Democracy, Human Rights and Intelligence Sharing

09 Sepper PUB_FINAL1 (Do Not Delete) 11/21/2010 5:33 PM
151
Democracy, Human Rights, and
Intelligence Sharing
ELIZABETH SEPPER*
ABSTRACT
In this Article, the author explores the networks used by intelligence agencies to
share intelligence and conduct joint operations with foreign counterparts worldwide.
Understanding how these intelligence networks operate, the author argues, is
imperative both for effective intelligence gathering and for a democratic society.
Characterized by secrecy, flexibility, and informality, intelligence sharing networks are
constrained almost exclusively by a shared professional ethos, rather than law.
According to the author, such an ethos can exert some degree of accountability to
professional norms, but has been strained by the inclusion of less professional and
often ruthless intelligence services in the network. Nonetheless, all such networks pose
serious threat to the preservation of liberal democracies in that they essentially govern
themselves. The very concept of democracy demands that an intelligence agency be
held accountable to a democratic body or officials outside of the agency itself. Yet, as
this Article shows, few democratically elected officials are aware of intelligence
sharing; and virtually no mechanism, other than self-regulation, provides oversight or
accountability for any intelligence agencys transnational activities. As a result,
through their network ties, intelligence agencies that are expected to serve democratic
interests have undermined foreign policy and circumvented safeguards established by
domestic law and international treaties. The author argues that this serious gap in the
rule of law must be filled and posits ways to render intelligence agencies more
accountable to the democracies they purport to serve.
SUMMARY
INTRODUCTION ............................................................................................................... 153

I. CHARACTERISTICS OF TRANSNATIONAL INTELLIGENCE SHARING
NETWORKS ........................................................................................................... 154
* Fellow, Columbia Law School and the Center for Reproductive Rights; LL.M. in International
Legal Studies, New York University School of Law; J.D., New York University School of Law; B.A., 2002,
Boston University. Special thanks to Stephen Holmes, who supervised my LL.M thesis upon which this
Article is based. Many thanks to Benedict Kingsbury, Simon Chesterman, Philip Alston, Samuel Rascoff,
and the participants of the Institute for International Law and Justice Scholars Workshop, particularly
Emily Berman and Jared Schott, for their comments and suggestions. My profound thanks to the editors
of the Texas International Law Journal for all their hard work.
152 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 46:151
A. Secret and Flexible Arrangements .............................................................. 156
B. Adherence to Professional Network Norms .............................................. 159
C. Reputational Sanctioning to Ensure Compliance with Professional
Norms ........................................................................................................... 163

II. CHALLENGES OF TRANSNATIONAL INTELLIGENCE COOPERATION FOR
DEMOCRACY AND HUMAN RIGHTS ................................................................. 166
A. Deficit of Democratic Accountability and Control ................................... 167
1. Lack of Transparency ........................................................................... 168
2. Absence of Democratic Oversight ...................................................... 169
B. Collusion to Avoid Democratically Determined Policies and Statutes ... 172
C. Undermining Foreign Policy and Human Rights through Connections
to Illiberal Agencies ..................................................................................... 174
1. Quid Pro Quo: Intelligence at What Price? ...................................... 175
2. Silencing Criticism and Bolstering Authoritarian Regimes ............. 176
3. Complicity of Liberal Democracies in Rights Violations ................. 178

III. A PROPOSAL FOR MORE PROFESSIONAL, DEMOCRATIC INTELLIGENCE
EXCHANGE .......................................................................................................... 182
A. Acculturation of Intelligence Agencies to Ethical Professional Norms ... 183
1. New Rules for Intelligence Cooperation ........................................... 183
a. Ethical Standards in Compliance with International Human
Rights Law ..................................................................................... 184
b. Caveats as a Requirement ............................................................. 185
c. Restrictions on the Use of Shared Information .......................... 187
2. Professionalization of Illiberal Agencies ............................................ 189
3. Limits on Quid Pro Quo ...................................................................... 192
B. Establishing Democratic Oversight and Accountability ........................... 193
1. Improve Transparency through Declassification .............................. 194
a. Declassification and Legislative Approval of Intelligence
Sharing Agreements ...................................................................... 194
b. Presumption in Favor of Declassification of Shared
Information .................................................................................... 195
2. Establish Democratic Control and Oversight .................................... 197
a. Set Clear Statutory Authority and Limitations on
Intelligence Cooperation .............................................................. 197
b. Develop Budgetary Control .......................................................... 200
C. Rely on and Bolster Law Enforcement Intelligence .................................. 201
1. Improve Law Enforcement Sharing Networks .................................. 202
2. Benefits of Law Enforcement Intelligence and the Involvement
of Courts ................................................................................................ 203

CONCLUSION ................................................................................................................... 206

2010] DEMOCRACY, HUMAN RIGHTS, AND INTELLIGENCE SHARING 153
INTRODUCTION
Although the global communityand the United States in particularhad long
discussed improving information sharing to combat transnational crime, the events of
September 11, 2001 brought networks for information sharing center-stage.
Coalition forces in Afghanistan formed networks to exchange information about
military operations; financial regulators established formal links to enforce sanctions;
and law enforcement agents institutionalized more frequent contact with other
services. Pride of place was reserved, however, for the intelligence agencies and their
networks.
The post-9/11 counterterrorism strategy was intelligence-driven, and
intelligence agencies and their networks would be at the helm. These agencies,
which had long networked with foreign counterparts, thus broadened and deepened
their connections. They incorporated new, and sometimes surprising, partners for
the exchange of information. Network exchanges became frequent and widespread
with ever more information shared with an increasing number of actors.
Almost ten years later, the role of intelligence agencies has not diminished. If
anything, their prominence has grown. Barely a week goes by without a government
touting its intelligence cooperation. Media reports hinting at the degree and
regularity of cooperation and intelligence sharing have become commonplace.
Yet, the specifics remain secret. For those outside the intelligence profession,
little is clear about how intelligence agencies cooperate with each other, which
partners are most involved, and how sharing mechanisms function in practice.
This Article argues that it is imperative that we understand these little known
intelligence networks, given their central role in global counterterrorism strategy and
their serious risk to democracy and accountability. Characterized by secrecy,
flexibility, and informality, the intelligence sharing networks are constrained almost
exclusively by a shared professional ethos, rather than law. Such an ethos can exert
some degree of accountability to professional norms, but has been strained by the
inclusion of less professional and often ruthless intelligence services in the network.
These networks, which essentially regulate themselves, pose an increasingly serious
threat to the preservation of liberal democracies. Intelligence agencies in a liberal
democracy are expected to act professionally, to inform foreign policy, and to serve
democratic interests. The very concept of democracy demands that an intelligence
agency be watched and held accountable by a democratic body or officials, outside of
the agency itself. Yet, as this Article shows, virtually no other mechanism provides
oversight or accountability for an intelligence agencys transnational activities.
Democratically elected officials, whether legislative or executive, often are entirely
ignorant of intelligence cooperation. With little oversight or regulation of their
network ties, agencies can circumvent domestic and international legal restraints and
collude with one another to the detriment of their respective states. These networks
thus strain our conceptions of the role of intelligence agencies, the effectiveness of
national control, and the democratic state itself. Recognizing that intelligence
cooperation and democratic principles will likely always be in tension, this Article
proposes several solutions to tip the balance from self-regulation to democratic
oversight and accountability.
Part I begins with an examination of currently known network arrangements
and their common characteristics. It explains that intelligence sharing networks
trade information on shared threats, function in secrecy, and operate with a high
level of informality. Most importantly, intelligence professionals identify with one
anotherespecially their trusted partners. Within their network arrangements,
agencies develop standards of professional behavior and enforce some degree of
compliance with them by sanctioning violations of professional standards.
Part II argues that in addition to the benefits they offer for the rapid exchange
of information, transnational intelligence network arrangements exact significant
costs for democracy and human rights. Without vigilant oversight, intelligence
networks have significant ability to undermine foreign policy and circumvent
safeguards established by domestic statutes and international treaties. The human
rights abuses committed by intelligence services in recent years are not isolated
examples but rather symptoms of a more systemic problem inherent in allowing
intelligence services to escape scrutiny of their transgovernmental ties. As Part II
will show, intelligence agencies through their network contacts have been able to
countermand policies made in democratic fora. Together, agencies have undermined
human rights protections, strengthened authoritarian governments, and outsourced
torture and abuse.
Accepting that intelligence networks can and should counter transnational
threats, Part III proposes several improvements to make intelligence sharing more
effective. It recommends two reinforcing methods to secure better, more democratic
and human rights-compliant intelligence exchange. First, agencies from liberal
democracies should establish more robust professional standards and use their
networks to acculturate less reputable agencies to more ethical and accurate
behavior. Second, democratic representatives and the global public should devote
greater time and attention to the problems that persist in transnational intelligence
networks. The transnational activities of agencies should be monitored and
regulated. Ultimately, intelligence agencies should be rendered accountable to the
democracies they purport to serve.
I. CHARACTERISTICS OF TRANSNATIONAL INTELLIGENCE SHARING
NETWORKS
Often presumed nationalistic and averse to cooperation, intelligence agencies in
fact have long networked with one another, sharing information and coordinating
operations to address mutual problems. Some contemporary intelligence networks
date back to the Cold War.
1
Many others materialized after the fall of the Soviet
Union, when agencies found that, in a globalized world, they were increasingly called
upon to combat problems spanning borders, such as drug and human trafficking,
organized crime, and terrorism.
2
After the attacks of 9/11, these networks were
1. See RICHARD ALDRICH, THE HIDDEN HAND: BRITAIN, AMERICA AND COLD WAR SECRET
INTELLIGENCE 89 (2001) (discussing the development of the Western intelligence community after
World War II). These intelligence exchange networks took on one of three forms: the client-server
network of the Union of Soviet Socialist Republics; Western, or Western-allied, agencies exchanging
information through NATO and more commonly bilateral arrangements; and former (or then) colonial
powers continuing ties with intelligence services in former colonies. Id. at 89, 400; Chris Clough, Quid
Pro Quo: The Challenges of International Strategic Intelligence Cooperation, 17 INTL J. OF INTELLIGENCE
& COUNTERINTELLIGENCE 601, 60304 (2004) (discussing the Soviet client-server network).
2. Richard J. Aldrich, Dangerous Liaisons: Post-September 11 Intelligence Alliances, HARV. INTL
REV., Fall 2002, at 50, 54 ([T]here is something to be said for viewing clandestine agencies and their
further expanded to include some unlikely and otherwise hostile intelligence partners
and, in some cases, deepened to allow more regular information exchange.
3

Although intelligence networks are often described as a one-on-one
relationship or as a hub and spokes configuration, they are best conceptualized as
a spider web with a multiplicity of connections expanding in every direction. The
complexity of connections cannot be overstated. A single intelligence agency may
have hundreds of ties and relationships to counterpart agencies worldwide. The
Canadian Security Intelligence Service (CSIS), for example, which has relatively
little independent ability to collect intelligence globally, has more than 250
information sharing arrangements with foreign security and intelligence
organizations.
4
Unlike the CSIS, the U.S. Central Intelligence Agency (CIA) has
global reach, including connections to more than 400 agencies.
5

These networks supply reach and competence far beyond that permitted by the
budgets and resources of each individual agency, giving members access to fiscal and
technological intelligence resources and global intelligence product.
6
Through
contacts with foreign intelligence agencies, an intelligence official receives
information she otherwise might not have because of her agencys limited technical,
geographical, human, or linguistic resources. Even the United Stateswhich has by
far the highest intelligence budget in the world, outstanding capability in technical
secret friendships as the product of structural change, most obviously of globalization.). Several factors
counseled in favor of wider cooperation. Regulation and enforcement at the national level proved
increasingly ineffective to contain transnational threats. Victor D. Cha, Globalization and the Study of
International Security, 37 J. OF PEACE RESEARCH 391, 394 (2000); Bjrn Mller-Wille, For Our Eyes
Only? Shaping an Intelligence Community Within the EU, Eur. Union Inst. for Sec. Studies Occasional
Paper No. 50, at 5 (Jan. 2004), available at http://www.iss.europa.eu/uploads/media/occ50.pdf (last visited
Sept. 30, 2010) (Detecting and assessing the so-called new threats correctly requires increased
intelligence cooperation between . . . agencies from different countries.). Technological advances
allowed previously impossible, rapid, real-time transmission of information, making intelligence sharing
more manageable. PETER GILL, ROUNDING UP THE USUAL SUSPECTS?: DEVELOPMENTS IN
CONTEMPORARY LAW ENFORCEMENT INTELLIGENCE 37 (2000); Andrew N. Liaropoulos, A (R)evolution
in Intelligence Affairs? In Search of a New Paradigm, Research Inst. for Eur. and Am. Studies Paper No.
100 14 (June 2006), available at http://www.isn.ethz.ch/isn/Current-Affairs/Security-Watch/Detail/?
size320=50&page320=3&ots591=0c54e3b3-1e9c-be1e-2c24-a6a8c7060233&lng=en&id= 31740.
3. Michael Herman, Ethics and Intelligence After September 2001, 19 INTELLIGENCE & NATL
SECURITY 342, 352 (2004). Intelligence exchange with counterparts in hostile nations may be
counterintuitive, but it is not uncommon. While only one state (or coalition of states) can win a war, trade
contract, or border dispute, most states benefit from reduced drug trafficking or terrorismit is not a zero
sum game.
4. Commn of Inquiry into the Actions of Canadian Officials in Relation to Maher Arar, A New
Review Mechanism for the RCMPs National Security Activities 140 (2006), available at http://www.sirc-
csars.gc.ca/pdfs/cm_arar_rcmpgrc-eng.pdf [hereinafter Arar Commission ReportA New Review
Mechanism]. See also Stphane Lefebvre, The Difficulties and Dilemmas of International Intelligence
Cooperation, 16 INTL J. OF INTELLIGENCE & COUNTERINTELLIGENCE 527, 534 (2003) (As of 2003,
ASIO [Australian Security Intelligence Organisation] had 233 liaison partners, distributed across 104
countries, and, as of 2002, CSIS had 230 liaison arrangements with agencies from 130 countries.).
5. Michael Herman, 11 September: Legitimizing Intelligence?, 16 INTL REL. 227, 233 (2002); Loch
Johnson, The Liaison Arrangements of the Central Intelligence Agency, in THE CENTRAL INTELLIGENCE
AGENCY: SECURITY UNDER SCRUTINY 85, 93 (Athan Theoharis et al. eds., 2006).
6. LOCH K. JOHNSON, SECRET AGENCIES: U.S. INTELLIGENCE IN A HOSTILE WORLD 122 (1996)
(Another means for less affluent nations to economize in their intelligence budgets is through sharing
arrangements with allies, commonly known as intelligence liaison or burden sharingan increasingly
attractive approach also for wealthy nations undergoing fiscal stress.).
areas, and world-wide reachdepends on foreign intelligence agencies with
comparative advantage in other specialized areas.
7

This Part provides a brief background of the defining characteristics of
transgovernmental intelligence networks. Such networks are transgovernmental in
that the relevant actors are not heads of state or foreign ministers, but rather
specialized government agencies.
8
Intelligence networks thus connect intelligence
agencies to their counterparts in other countries for the purpose of exchanging
information.
9
The primary characteristics of these networks, as Section A describes,
are absolute secrecy and flexibility in the form, style, and mechanisms of sharing.
Section B explains that networks give rise to a professional community that sets its
own standards and norms. Within this community, the extent to which an
intelligence agency receives valuable information is decided by its reputation (the
degree to which it can be trusted and add value). Therefore, as Section C argues,
reputation is sacrosanct. As a result, in the absence of control by more democratic
organs, networks can exert some compliance with professional norms through
sanctioning.
A. Secret and Flexible Arrangements
Although transgovernmental networks generally lack transparency as compared
to other institutions, intelligence sharing networks operate with the highest levels of
secrecy. The very structures through which agencies share information are among
the intelligence communitys top secrets.
10
Even where the existence of such
7. See, e.g., SHLOMO SHPIRO, THE COMMUNICATION OF MUTUAL SECURITY: FRAMEWORKS FOR
EUROPEAN-MEDITERRANEAN INTELLIGENCE SHARING 6 (2001), available at www.nato.int/acad/fellow/
99-01/shpiro.pdf (noting that most intelligence services of Mediterranean countries do not have global
coverage but possess the capabilities to train and operate agents throughout the region in a better way
than the US intelligence community or many of their European counterparts).
8. Kal Raustalia, The Architecture of International Cooperation: Transgovernmental Networks and
the Future of International Law, 43 VA. J. INTL L. 1, 3 (2002).
9. A number of scholars and intelligence professionals define intelligence functionally, as information
generated by intelligence agencies. See MICHAEL HERMAN, INTELLIGENCE POWER IN WAR AND PEACE
96 (1996) (stating that [u]nprocessed single-source data is officially information, and becomes
intelligence only through the conversion of information into intelligence through collation, evaluation,
analysis, integration and interpretation); Roy Godson, Intelligence: An American View, in BRITISH AND
AMERICAN APPROACHES TO INTELLIGENCE 3 (K.G. Robertson ed., 1987) (noting that [g]overnments
and even intelligence services rarely define intelligence explicitly. Instead, they develop policies,
programmes, and patterns of organization. These demonstrate the role the state visualizes for itself, and
its concepti.e. definitionof intelligence.); Mller-Wille, supra note 2, at 7. For the purpose of this
Article, intelligence is defined functionally as information generated by intelligence agencies. This Article
will therefore use the terms intelligence and information interchangeably. The definition of intelligence
used here excludes open source intelligence, which is publicly available, since it generally does not raise
the human rights and accountability issues endemic to other forms of intelligence.
10. Martin Rudner, Hunters and Gatherers: The Intelligence Coalition Against Islamic Terrorism, 17
INTL J. INTELLIGENCE & COUNTERINTELLIGENCE 193, 222 (2004) [hereinafter Rudner, Hunters and
Gatherers]; Aldrich, supra note 2, at 50 (Intelligence alliances are among the most closely guarded secrets
of clandestine agencies.). In the United States, the fact of cooperation and details of intelligence sharing
arrangements are kept classified as a general rule. DIRECTOR OF CENTRAL INTELLIGENCE DIRECTIVE
1/10, SECURITY CLASSIFICATION GUIDANCE ON LIAISON RELATIONSHIPS WITH FOREIGN INTELLIGENCE
ORGANIZATIONS AND FOREIGN SECURITY SERVICES (effective Dec. 14, 1982), available at
http://www.fas.org/irp/offdocs/dcid1-10.htm.
networks has been revealed, the essential elementsthe participants, methods of
operation, and agreements themselvesremain classified.
11

In the rare case, these secret intelligence networks have been negotiated and
agreed to by the heads of states, likely resulting in executive agreements between
states.
12
Such is the case, for instance, with the most formalized and best-known
intelligence network between the signals intelligence (SIGINT) services of the
United Kingdom, Canada, Australia, New Zealand, and the United States.
13
The
allied intelligence superpowers emerging from World War II negotiated the United
Kingdom-United States Intelligence Agreement (UKUSA Agreement) in 1947 at
a high level of government to achieve global SIGINT coverage, which neither could
achieve alone.
14
Canada, Australia, and New Zealand rapidly signed on as second
parties to the agreement. Unlike more informal arrangements, this agreement
institutionalized and divided the collection and exchange of SIGINT between all five
agenciesresulting in the most integrated known intelligence network.
15
These
agencies were namely the U.S. National Security Agency (NSA), U.K. Government
Communications Headquarters (GCHQ), Australian Defence Signals Directorate
(DSD), Canadian Communications Security Establishment (CSE), and New Zealand
Government Communications Security Bureau (GCSB).
While most academic studies have focused on the UKUSA Agreement, it
represents an aberration both in its formality and its degree of integration. Other
multilateral networks have developed more informally. Best known are the
numerous, long-standing plurilateral clubs of European agencies focused on
intelligence exchange against transnational threats.
16
TRIDENT, the cooperative
11. The UKUSA Agreement set the precedent for institutionalizing the requirement of absolute
secrecy as to the agreements existence. JEFFREY T. RICHELSON, THE U.S. INTELLIGENCE COMMUNITY
294 (4th ed., 1999) (1985) (Despite numerous references to the agreements in print, officials of some of
the participating governments have refused to confirm not only the details of the agreement but even its
existence.). While recently there has been some public discussion of UKUSAs existence, freedom of
information requests to relevant governments generate no information. Kevin J. Lawner, Post-Sept. 11
th

International Surveillance ActivityA Failure of Intelligence: The ECHELON Interception System & the
Fundamental Right to Privacy in Europe, 14 PACE INTL L. REV. 435, 44445 (2002).
12. Rudner, Hunters and Gatherers, supra note 10, at 195. The prime example involves the early
SIGINT sharing agreements to which the Commonwealth countries are signatories. JEFFREY T.
RICHELSON & DESMOND BALL, THE TIES THAT BIND: INTELLIGENCE COOPERATION BETWEEN THE
UKUSA COUNTRIES 47 (1985). See also Matthew M. Aid, The National Security Agency and the Cold
War, in SECRETS OF SIGNALS INTELLIGENCE DURING THE COLD WAR AND BEYOND 27, 33 (Matthew M.
Aid & Cees Wiebes, eds., 2001). Canada even permitted the United Kingdom to represent its interests in
negotiations. Rudner, Hunters and Gatherers, supra note 10, at 196; Martin Rudner, Britain Betwixt and
Between: UK SIGINT Alliance Strategys Transatlantic and European Connections, 19 INTELLIGENCE &
NATL SECURITY 571, 574 (2004) [hereinafter Rudner, Britain Betwixt and Between] (Later, countries
like Denmark, Germany, and Turkey were reportedly included in a somewhat looser, more limited
association as so-called Third Parties, usually by virtue of bilateral arrangements . . . .).
13. David Alan Jordan, Decrypting the Fourth Amendment: Warrantless NSA Surveillance and the
Enhanced Expectation of Privacy Provided by Encrypted Voice Over Internet Protocol, 47 B.C. L. REV.
505, 510 (2006) (defining SIGINT as information intercepted by technological means including
communications intelligence, electronics intelligence, and foreign instrumentation signals intelligence).
14. SHPIRO, supra note 7, at 12.
15. RICHELSON, supra note 11, at 293.
16. Rudner, Hunters and Gatherers, supra note 10, at 210 (The Club of Berne, dating from 1971, is a
forum for the heads of the security services of European Union (EU) member countries . . . . The Club
has no statutory mandate; neither does it report to any authority within the EU framework.).
Established in the 1970s, the Kilowatt group now unites twenty-four statesincluding most EU members
arrangement between intelligence services of Israel, Turkey, and Ethiopia, is another
example, although little is known about it.
17
Informal regional arrangements,
supplemented by bilateral ties, have been set up throughout the Asia-Pacific region
as well.
18

The great majority of intelligence is shared, however, not through formal or
multilateral agreements but rather through informal, typically bilateral network
arrangements.
19
Even among the five UKUSA countries, the UKUSA Agreement
does not govern all intelligence exchange. Instead, hundreds of less formal ties
connect the various agencies of the five countries.
20

Most commonly, national intelligence agencies (rather than governments)
negotiate memoranda of understanding (MOUs), setting out modalities of
intelligence exchange.
21
In contrast to treaties, or even executive agreements, MOUs
do not require approval by national legislators (or foreign policy ministers) and can
be implemented by the agencies themselves.
22
As non-binding, soft law agreements,
MOUs regularize contacts and cooperation between individual intelligence services.
23

Considered flexible and easy to establish, they create a loose and adaptable
framework in which to share information, ideas, and resources.
24
These written
agreements establish both the degree and method of cooperation.
25

Less formal arrangements based on oral agreements or personal friendships are
also widespread among intelligence agents. Generally speaking, they govern
information exchanges in the absence of an MOU, supplant existing MOUs, and
characterize ad hoc contacts during crises.
26
Frequently operating below the level of
and Canada, Israel, Britain, the United States, and South Africa. Lefebvre, supra note 4, at 531.
17. SHPIRO, supra note 7, at 15.
18. See, e.g., Simon S.C. Tay, Asia and the United States After 9/11: Primacy and Partnership in the
Pacific, 28 FLETCHER F. WORLD AFF. 113, 118 (2004) (ASEAN signed an anti-terrorism pact that would
commit members to cooperate in stemming the flow of funds to terrorist groups as well as sharing
intelligence and increasing police cooperation in order to prevent, disrupt and combat international
terrorism.); ANDREW SCHEINESON, COUNCIL ON FOREIGN REL., PUB. NO. 10883, THE SHANGHAI
COOPERATION ORGANIZATION (Mar. 24, 2009), available at http://www.cfr.org/publication/10883/ (noting
that under the Shanghai Cooperation Organization, the security services of Russia, China, Kazakhstan,
Kyrgyzstan, Tajikistan, and Uzbekistan share intelligence and conduct joint counterterrorism drills).
19. HERMAN, supra note 9, at 207.
20. RICHELSON & BALL, supra note 12, at 141. To give an indication of the sheer number of agencies
involved, the United States alone has fifteen intelligence agencies. Norman C. Bay, Executive Power and
the War on Terror, 83 DENV. U. L. REV. 335, 350 (2005).
21. Lefebvre, supra note 4, at 533 (Bilateral liaison arrangements are a defining characteristic of the
intelligence world. Set up formally (i.e., with the signing of a Memorandum of Understanding) or
informally (on the basis of an unwritten, gentlemanly agreement), they pay particular attention to the
participants protection of their intelligence.).
22. Anne-Marie Slaughter, The Accountability of Government Networks, 8 IND. J. GLOBAL LEGAL
STUD. 347, 359 (2001).
23. Sol Picciotto, Networks in International Economic Integration: Fragmented States and the
Dilemmas of Neo-Liberalism, 17 NW. J. INTL L. & BUS. 1014, 1047 (1997); Raustalia, supra note 8, at 22.
24. Raustalia, supra note 8, at 22. The relative level of formality or informality is not, however,
determinative of the networks efficiency. GILL, supra note 2, at 36.
25. See SHPIRO, supra note 7, at 13 (discussing three interrelated levels of cooperationmacro,
micro, and mesoand what each level entails).
26. Id. See also Commn of Inquiry into the Actions of Canadian Officials in Relation to Maher Arar
Policy Review, National Security and Rights and Freedoms: A Background Paper to the Commissions
Consultation Paper 16 (Dec. 10, 2004), available at http://epe.lac-bac.gc.ca/100/206/301/pco-
bcp/commissions/maher_arar/07-09-13/www.ararcommission.ca/eng/National%20Security%20and%20
official control, informal cooperative arrangements also allow contact even when
interaction with a certain intelligence agency (or state) is officially disfavored.
27

Ultimately, these varied networks share numerous characteristics. We can
expect them to form against common threats, to remain secret, and to operate with a
high level of informality. Most importantly, intelligence professionals exhibit a high
level of identity with one anotherespecially their trusted partners. This
professional community and its potential for sanctioning violations of professional
standards will be explored further in Section B.
B. Adherence to Professional Network Norms
Through their network ties, intelligence agencies form what Peter Haas terms
an epistemic community, that is, a network of professionals who share a
specialized expertise and knowledge in a particular field.
28
Shared practices,
normative principles, and evaluative criteria within their domain give agencies a
sense of identity with their counterparts in other countries.
29
These commonalities
make it possible, indeed probable, that agencies network with one another.
30

Three principal attributes are common to all intelligence services. First among
these is the requirement of secrecy in performing collection, analysis,
counterintelligence, and covert operations.
31
Second, security clearances and access
to classified information generate a deliberate culture of identification as a member
of a unique, even elite club.
32
The secrecy of intelligence agencies magnifies this
corporate cohesivenesswhich all professions exhibit to some degreegenerating a
sense of superiority and an affinity for fellow intelligence professionals.
33
Third, all
intelligence agencies understand themselves to assist national-level decision-making
and serve in the defense of the state.
34
They consequently, at least among
professionalized services, attempt to act in accordance with a professional ethos of
objectivity and pursuit of truth so that decisions can be made on valid, reliable
information.
35

Rights%20and%20Freedoms.pdf [hereinafter National Security Background Paper to Arar Commission]
(commenting on the impact of international cooperation during national security investigations).
27. GILL, supra note 2, at 36 (observing that with regard to law enforcement intelligence, [s]ome of
these contacts will be legitimated via official agreements while others will be officially frowned on).
28. Peter M. Haas, Introduction: Epistemic Communities and International Policy Coordination, 46
INTL ORG. 1, 3 (1992).
29. Id.
30. Id.
31. Thomas C. Bruneau & Kenneth R. Dombroski, Reforming Intelligence: The Challenge of Control
in New Democracies, in WHO GUARDS THE GUARDIANS AND HOW: DEMOCRATIC CIVIL-MILITARY
RELATIONS 157, 165 (Thomas C. Bruneau & Scott D. Tollefson eds., 2006).
32. Id.
33. Id. See also HERMAN, supra note 9, at 327 (observing a greater-than-usual sense of difference
from other walks of life).
34. Bruneau & Dombroski, supra note 31, at 166.
35. LOCH K. JOHNSON, BOMBS, BUGS, DRUGS, AND THUGS: AMERICAS QUEST FOR SECURITY 100
(2000) (The ethosin theory at least and usually in practiceis objectivity, and the goal is to provide
decision makers with accurate, timely, and relevant information and insight . . . .); accord Herman, supra
note 3, at 345.
Within this community of networked intelligence agencies, intelligence
producers develop standards and normsmore or less formally articulatedto
which members must adhere.
36
The professional standards of the UKUSA network
in particular are so well-developed that, according to James Bamford, an expert on
the NSA, the UKUSA community constitutes a unique supranational body,
complete with its own laws, oaths, and language, all hidden from public view.
37

What then are the network standards with which intelligence agencies must comply,
or stand accountable before their peers?
Information security and handling, as well as provision of reliable information,
figure prominently in these standards. Because every exchanged secret risks
disclosing sources and methods and thereby jeopardizing future information,
38

sending agencies require the enforcement of tight rules for the handing and storage
of shared intelligence. Communications, physical facilities, and personnel must be
secure and able to protect sensitive information.
Two principal rules ensure that shared intelligence will be kept confidential and
not further disseminated without the sending states consent. First, the need to
know principle, typical of intelligence agencies worldwide, dictates that individuals
only receive access to shared, classified information when they need it for their work.
Within intelligence networks, a sending agency may stipulate that only it or the
specific addressee of the intelligence may make the decision that an individual
needs to know.
39
In some cases, shared intelligence may even be designated only
for the specific agency with which it is shared; the latter being expected to restrict
circulation even with its sister agencies.
40

Second, information shared remains the property of the originator of the
information. Known as originator or user control, this rule restricts subsequent
dissemination of information.
41
Before confidential information is released, a
36. See Simon Chesterman, The Spy Who Came in from the Cold War: Intelligence and International
Law, 27 MICH. J. OF INTL L. 1071, 1099 (2007) (arguing that intelligence agencies interactions have
generated some agreed-upon normative guidelines and a professional ethic that suggests a kind of
community that generates, adapts, and internalizes rules).
37. JAMES BAMFORD, BODY OF SECRETS: THE ANATOMY OF THE ULTRA-SECRET NATIONAL
SECURITY AGENCY: FROM THE COLD WAR THROUGH THE DAWN OF A NEW CENTURY 40304 (2001);
RICHELSON & BALL, supra note 12, at 5 (stating that the Agreement includes access to intelligence and
security arrangements, standardized codewords, security agreements that all employees of each UKUSA
agency must sign, and procedures for storing and disseminating material); James I. Walsh, Intelligence-
Sharing in the European Union: Institutions Are Not Enough, 44 J. COMMON MARKET STUD. 625, 630
(2006) (It establishes common security procedures and standardized technical terms, code words and
training across the participating countries intelligence services, ensuring that shared intelligence is
handled in a consistent manner and is unlikely to be misinterpreted by a receiving state.).
38. Mller-Wille, supra note 2, at 15 (All intelligence collectors are concerned about the security of
their sources and their methods of collecting information. If these are uncovered, access to the
information will be jeopardized. In addition, they may want to protect the information itself.).
39. Alasdair Roberts, Entangling Alliances: NATOs Security of Information Policy and the
Entrenchment of State Secrecy, 36 CORNELL INTL L.J. 329, 33738 (2003).
40. Commn of Inquiry into the Actions of Canadian Officials in Relation to Maher Arar: Policy
Review, Accountability and Transparency: A Background Paper to the Commissions Consultation Paper 7
(Dec. 10, 2004) [hereinafter Accountability and Transparency Background Paper to Arar Commission].
41. Roberts, supra note 39, at 338 (The principle was central to the agreement signed by NATO
members in January 1950.). Intelligence sharing between domestic agencies of the same national
intelligence community also occasionally uses a classification caveat, known as ORCON (originator
controlled). The CIA uses ORCON most extensively than any other U.S. agencyperhaps because it
receives much intelligence from foreign agencies. MICHAEL A. TURNER, WHY SECRET INTELLIGENCE
receiving agency is obliged to guarantee that it will not be disclosed.
42
The provider
also sets the classification level, which the recipient is bound to respect, and reserves
the right to maintain or change it.
43
Sometimes a sending agency authorizes the
future sharing of information with the services of certain other states a priori; at
other times, it only approves further distribution on an ad hoc basis.
44
This is to avoid
what is known as a Trojan Horsewhere a receiver shares the intelligence with
third parties that have not obtained the providers security clearance.
45

To avoid intelligence failure caused by inaccurate intelligence, professional best
practices require screening exchanged intelligence for relevance and attaching
written caveats to indicate the informations reliability.
46
These caveats can allow a
receiving agency to judge a piece of informations probative value. If the intelligence
has not been tested, the receiving agency should not rely on it without corroboration.
Professional standards provide that, in all cases, an agency generally should test or
corroborate received intelligence through its own independent collection capabilities.
Alternately, services can cross-check the information through contacts with other
foreign counterparts.
47
For example, if France provides Germany with intelligence
indicating that a specific individual is associated with terrorist activities and crossed
into Germany in a given month, German intelligence might request information
regarding that person from intelligence agencies in bordering states. If other
information proved that he or she had been in Singapore for the entirety of that
month, Frances shared intelligence would be deemed incorrect.
48

FAILS 9495 (2006).
42. Accountability and Transparency Background Paper to Arar Commission, supra note 41, at 7
(noting that much of the intelligence Canada receives is designated as confidential and released only on
the guarantee that it will not be publicly revealed).
43. Simon Duke, Intelligence, Security and Information Flows in CFSP, 21 INTELLIGENCE & NATL
SECURITY 604, 611 (2006).
44. Id. (noting senders may stipulate, for example, that information may be shared within NATO or
members of the EUs Partnership for Peace framework who have adequate, pre-approved security
agreements).
45. Mller-Wille, supra note 2, at n.21 (setting out reasons a receiving agency might pass on
information to a third agency).
46. Commn of Inquiry into the Actions of Canadian Officials in Relation to Maher Arar, Report of
the Events Relating to Maher Arar: Analysis and Recommendations 2223 (2006), available at
http://www.sirc-csars.gc.ca/pdfs/cm_arar_rec-eng.pdf [hereinafter Arar Commission ReportAnalysis and
Recommendations]. With regard to intelligence delivered to the U.S. CIA and DIA in the lead-up to the
Iraq war, foreign services generally caveated intelligence properly. Report of the Select Committee on
Intelligence on the Use by the Intelligence Community of Information Provided by the Iraqi National
Congress, 109th Cong., 2d Sess. 59, 16768 (Sept. 8, 2006) [hereinafter Senate Report on Iraqi National
Congress]. However, some less dependable intelligence agencies may have fed reports from the Iraqi
National Congress without identifying their source, therefore causing general unreliability and
contributing to intelligence failure. Id. at 5.
47. Paradoxically, for these agencies, the only way the quality of intelligence exchanged can be
improved is through further sharing. Mller-Wille, supra note 2, at 16 (The accuracy can only be verified
or falsified if the collection of intelligence is increased, not if it is reduced.).
48. A disadvantage of this approach lies in the potential for circular reporting, where [a] snippet
of data may be passed in a full circle, and thus, when received by the originating nation, may be considered
corroboration for what might have been only a tentative assumption in the first place. Clough, supra
note 1, at 606. In our example, if Germany had received the same intelligence from the United States as
from France, it might accept that as corroborationeven if the U.S. had simply passed on intelligence
received from French intelligence. For examples of networked connections that led to circular reporting,
see BAMFORD, supra note 37, at 417.
Between closer network partners, standards may also include agreements to
distribute all intelligence gathered on a certain issue or by a specific collection
discipline (such as SIGINT), or to refrain from intercepting communications of a
partners nationals or using the network to gather economic information for national
corporate interests.
A reputation for compliance with these norms helps determine the volume,
type, and quality of information an agency receives and thus correlates to the
networks overall level of efficiency.
49
Highly trusted partners receive intelligence in
real-time and on a frequent basis, and may share intelligence gathering resources
such as satellites or listening stations.
50
At one extreme, the ECHELON system
between the UKUSA SIGINT agencies represents a tightly integratedand likely
unprecedentednetwork for sharing intelligence, spanning collection centers
worldwide and enabling the agencies to collect an enormous volume of
communications, as many as three billion a day.
51
According to Nicky Hager, who
wrote the first detailed expos of the system, the network allows the stations of each
partner agency to function as components of an integrated whole.
52
The example
of ECHELON shows that networks function mostly efficiently and effectively when
composed of partners that can be trusted to abide by professional standards.
In order for an intelligence agency to gain trust within the network, it must
establish a good professional reputation.
53
Because networks facilitate transmission
of information relating to a members reputation for competence and
trustworthiness,
54
intelligence officials are familiar with their peers reputations,
which vary even among agencies in the same country.
55
Many intelligence agencies
are acutely aware that their performance will be measured by their counterparts
49. Wenpin Tsai & Sumatra Ghoshal, Social Capital and Value Creation: The Role of Intrafirm
Networks, 41 ACADEMY OF MANAGEMENT J. 464, 467 (2004). Chris Clough identifies five levels of
intelligence sharing: (1) raw intelligence product revealing the source and product in detail; (2) all or
part of raw product without exposing the source; (3) summary of data (4) finished analysis of data; and
(5) policy conclusions resulting from the intelligence. Clough, supra note 48, at 603.
50. See Clough, supra note 1, at 603605 (noting examples of close intelligence cooperation between
countries).
51. BAMFORD, supra note 37, at 404. The partners share a massive computer network, linking
computer systems belonging to members worldwide and hosted at NSA headquarters. Id. Frances
SIGINT agency also has similar partnerships, sharing satellites with Italian, Spanish, Belgian, and German
intelligence. Duke, supra note 43, at 623; The New Challenges Facing European IntelligenceReply to
the Annual Report of the Council, Report Submitted on Behalf of the Defence Committee, Assembly of
Western European Union Doc. A/1775, paras. 7780 (June 4, 2002) [hereinafter New Challenges].
52. NICKY HAGER, SECRET POWER: NEW ZEALANDS ROLE IN THE INTERNATIONAL SPY
NETWORK 29 (1996). Nicky Hagers report is considered among the best informed analyses, although the
existence of the ECHELON system have never been officially confirmed. Rudner, Britain Betwixt and
Between, supra note 12, at n.31. See also Lawrence D. Sloan, Note, Echelon and the Legal Restraints on
Signals Intelligence: A Need for Reevaluation, 50 DUKE L.J. 1467, 1470 (2001) (The closest a
representative of the United States intelligence community has come to publicly confirming the existence
of ECHELON was when the Director of Central Intelligence, George Tenet, referred to the so-called
ECHELON program of the National Security Agency in congressional testimony.).
53. ANNE-MARIE SLAUGHTER, A NEW WORLD ORDER 49, 5455 (2004) (citing Majone who
referred to networks in the European Union as bearers of reputation).
54. Id. at 54 (noting that evident violations of those norms would quickly be transmitted across the
network, raising the cost of those violations).
55. Stephen Fidler, The Human Factor: All Is Not Well in Clandestine Intelligence Collection, FIN.
TIMES (LONDON), July 7, 2004, at 15.
domestically and abroad.
56
As former intelligence official Michael Herman says, [i]n
secret intelligence more than in most activities, a good reputation is slowly gained,
and easily lost.
57

C. Reputational Sanctioning to Ensure Compliance with Professional Norms
Whereas numerous civilian controls, including statutory limits, budgetary
oversight, and professional rewards, govern actions of law enforcement and military
officials, intelligence professionals even in democratic states generally rely on
professional norms to delineate the boundaries of their own transnational actions
and to hold partners accountable.
58
The written agreements (whether executive
agreements or MOUs) and oral understandings that apply to intelligence sharing
networks are not legally enforceable; there are no judicial remedies or political
sanctions to provide redress for breaches.
59
Nonetheless, networks can create
conditions under which professional standards come to be respected, despite the lack
of legal sanctions.
Within networks, real and imagined pressures from the intelligence peer group
drive compliance with professional norms in a process Ryan Goodman and Derek
Jinks have labeled acculturation.
60
These pressures include (1) the imposition of
social-psychological costs through shaming or shunning and (2) the conferral of
social-psychological benefits through displays of public approval.
61
Goodman and
Jinks suggest this drive toward acculturation may actually work better within a
denser network of relationships, because the network partners know one another,
have multiple interactions, and seek to remain members in good standing.
62

Unlike persuasion, acculturation does not require that actors actively assess
the content of a particular messagea norm, practice, or beliefand change their
56. See HERMAN, supra note 9, at 207 (explaining that most intelligence agencies are producing
partly for an international audience of fellow-professionals, as well as their primary national recipients).
57. Id.; Stephen Fidler & Mark Huband, A Special Relationship? The US and UK Spying Alliance Is
Put Under the Spotlight, FIN. TIMES (London), July 6, 2004, at 17.
58. Bruneau & Dombroski, supra note 31, at 164. Arthur S. Hulnick & Daniel W. Mattausch, Ethics
and Morality in United States Secret Intelligence, 12 HARV. J. L. & PUB. POLY 509, 509, 520 (1989) (Ethics
are often defined as behavior relating to professional standards of conduct. As in any other profession,
such standards exist in the field of intelligence, even if these standards require behavior that is
unacceptable for private citizens.); SLAUGHTER, supra note 53, at 54. See also Harold Hongju Koh,
Transnational Legal Process, 75 NEB. L. REV. 181, 204 (1996) (As transnational actors interact, they
create patterns of behavior and generate norms of external conduct which they in turn internalize.).
59. See Ruth W. Grant & Robert O. Keohane, Accountability and Abuses of Power in World Politics,
99 AM. POL. SCI. REV. 29, 35 (2005) (stating that one can expect accountability to operate chiefly through
reputation and peer pressure, rather than in more formal ways).
60. Harold Hongju Koh, How is International Human Rights Law Enforced?, 74 IND. L.J. 1397, 1404
(1999) (discussing his theory of transnational legal process, which explains legal compliance at the state
level in a way that bears numerous similarities to Jinks and Goodmans theory of acculturation); see also
RYAN GOODMAN & DEREK JINKS, SOCIALIZING STATES: PROMOTING HUMAN RIGHTS THROUGH
INTERNATIONAL LAW 14 (forthcoming 2011), available at http://www.lcil.cam.ac.uk/Media/lectures/pdf/
GoodmanJinksSocializingStatesMay2009.pdf ([C]onnectedness with a reference group determines
conformity to a social norm as opposed to the substantive content of a norm.).
61. GOODMAN & JINKS, supra note 60, at 8.
62. Id. at 56.
minds.
63
For example, if Jordanian intelligence understands that complying with
standards like caveating for reliability or not sharing information obtained through
torture is a precondition for inclusion and good reputation in the network, it will
obey the standards out of a desire to join or become a respected member of the
group.
64
The agency need not have an inherent belief in the normative legitimacy of
the rule to comply. It may even consider torture to be a useful interrogation method,
but desist from the practice for the purpose of information sharing because of its
desire to be an esteemed network partner. Classic social network analysis suggests
that the agencythrough network interactionsmay then come to internalize those
norms, resulting in a higher rate of compliance and greater incentives to maintain a
good reputation.
65

Experience shows that when prestigious actors within the intelligence networks
transmit network standards and exert pressure on their peers, other intelligence
agencies comply even when not admitting the legitimacy of the rule. The United
States and the United Kingdom, in particular, have acculturated their partners to
relatively arbitrary standards of behavior. At the insistence of U.S. intelligence, for
example, positive vetting of intelligence recruits for disloyalty or allegiance to radical
groups remains the professional standard
66
despite its dubious contribution to
increasing security. Even the United States closest allies must adopt such
regulations in order to secure its cooperation.
67
In a similar pattern, the U.S. Drug
Enforcement Agency has used its cooperative networks to acculturate partners to
drug trafficking detection, investigation, and interdiction; as a result, methods of
agencies are now very similar.
68

63. Id. at 34, 5 (The touchstone of the overall process [of persuasion] is that actors are consciously
convinced of the truth, validity, or appropriateness of a norm, belief, or practice.).
64. Id. at 7 (arguing that compliance can be motivated by the social-psychological benefits of
conforming to group norms and expectations (such as the cognitive comfort associated with both high
social status and membership in a perceived in-group)).
65. See GOODMAN & JINKS, supra note 60, at 8 (discussing the various social pressures that cause
internalization of and conformation with group behavioral patterns). Initially, however, those less-
professionalized services that join the network may abide by network standards out of pure self-interest,
rather than any internal sense of normativity. Id. at 89.
66. SHPIRO, supra note 7, at 3940 (explaining that development of intelligence cooperation
frameworks would depend on developing effective systems of information classification acceptable on
both cooperating sides. It also requires NATO authorities to accept, or at least endorse, the personnel
vetting practices and security clearances of potential cooperation partner intelligence services.). See also
SLAUGHTER, supra note 53, at 172 (observing a similar phenomenon with regards to MOUs concluded by
the U.S. Securities and Exchange Commission and foreign regulators implementation of U.S.-style
regulation).
67. Aldrich, supra note 2, at 425 (Soon after U.S.-U.K. peacetime intelligence sharing began, the
advisor said . . . Officials have already offered the procedure now proposed [positive vetting] and nothing
short of that offer . . . will secure their cooperation.).
68. GILL, supra note 2, at 75 (Americanization rather than harmonization fairly describes global
evolution of law enforcement since the 1960s, especially regarding drugs.). Foreign drug enforcement
agencies are not necessarily persuaded of the value of these techniques, rightly so since international drug
enforcement efforts have generally failed. Diane Marie Amann, The Rights of the Accused in a Global
Enforcement Arena, 6 ILSA J. INTL & COMP. L. 555, 556 (2000). However, drug or transnational crime
intelligence units in illiberal countries have improved and adopted more professional methods in spite of
the otherwise repressive tactics typically employed by other intelligence units. See, e.g., Seth G. Jones et
al., Natl Security Research Division, Securing Tyrants or Fostering Reform? U.S. Internal Security
Assistance to Repressive and Transitioning Regimes, RAND 85 (2006), available at http://www.rand.org/
pubs/monographs/2006/RAND_MG550.pdf (noting that Uzbek Sensitive Investigation Unit, charged with
countering drug trafficking, now operates within Western norms with the support of the U.S. DEA).
To counteract incentives to depart from professional standards, intelligence
networks rely heavily on reputational sanctioning. Anecdotal evidence suggests that
professionalized services agonize over their foreign partners potential disapproval.
The British government, for example, sought an injunction against Spycatcher, an
expos of British intelligence, because it would cause damage to the reputation of
the services in the eyes of foreign allies.
69
For intelligence officers, the
consequences of exceeding [international professional norms is] unacceptable
personally and professionally, nationally and internationally.
70

Exclusion from a network because of a bad reputation is improbablelargely
because most partners need the comparative advantage that the other provides.
Nevertheless, without a good reputation, the individual agencies (or agents) may find
themselves deprived of necessary information. For example, in the early 1960s, the
relationship of the British Secret Intelligence Service (MI6) with the CIA fell to a
low because security lapses and molehunts damaged the CIAs trust in MI6. MI6
remained valuable, however, due to high-quality intelligence it gathered from a few
select sources behind the Iron Curtain.
71
The history of Communist agent
penetration of the German Federal Intelligence Service (or Bundesnachrichtendienst
(BND)) resulted in its allies withholding the most sensitive or high-resolution
intelligence.
72
Even today, residual fears of Russian infiltration sometimes cause
Germanys allies to hold back intelligence.
73
Although trilateral negotiations
between the BND, NSA, and GCHQ resulted in a U.S. eavesdropping station jointly
manned by BND and NSA officials, the BND is not allowed access to
communications gathered by the NSA.
74
Those services that develop a reputation for
providing bad intelligence to the network likewise may find their flow of shared
intelligence reduced to a trickle.
Misuse of intelligence may also reduce sharing. For example, connections
between the CIA and Mossad were interrupted at least twice. The first time, Israel
used U.S. intelligence to bomb Iraqs nuclear reactor in 1981 over the CIAs
objections; the second time, the Mossad was revealed to have a spy within the U.S.
69. PATRICK BIRKINSHAW, FREEDOM OF INFORMATION: THE LAW, THE PRACTICE AND THE IDEAL
38 (3rd ed. 2001); see also Kent Pakel, Integrity, Ethics and the CIA: The Need for Improvement, STUDIES
IN INTELLIGENCE 85 (Spring 1998) (stating that in interviews, many intelligence officials at the CIA
described a tyranny of reputation, in which a bad call can stay with you for three years).
70. See Chesterman, supra note 36, at 1097 (arguing that this notion rings true). See also Anne-
Marie Slaughter & William Burke-White, The Future of International Law is Domestic (or, the European
Way of Law), 47 HARV. INTL L.J. 327, 334 (2006) (They build trust and establish relationships among
their participants that create incentives to establish a good reputation and avoid a bad one.).
71. STEPHEN DORRIL, MI6: INSIDE THE COVERT WORLD OF HER MAJESTYS SECRET
INTELLIGENCE SERVICE 703 (2000). See also RICHELSON & BALL, supra note 12, at 6 (noting that
security lapses have resulted in compromise of UKUSA intelligence numerous times).
72. See James Risen, Bonn Sniffs for Russian Moles, Worrying C.I.A., N.Y. TIMES, June 4, 1998 (One
U.S. intelligence officer asserts that the BND has had a history of penetration, and the truth is we have
never really taken them too seriously as an intelligence organization).
73. Charles Grant, Intimate Relations: Can Britain Play a Leading Role in European Defenceand
Keep Its Special Links to US Intelligence? 7 (Ctr. for Eur. Reform Working Paper Apr. 2000), available at
http://www.cer.org.uk/pdf/cerwp4.pdf.
74. Erich Schmidt-Eenboom, The Bundesnachrichtendienst, the Bundeswehr and Sigint in the Cold
War and After, in SECRETS OF SIGNALS INTELLIGENCE DURING THE COLD WAR AND BEYOND 129, 147
50 (Matthew M. Aid & Cees Wiebes eds., 2001); DORRIL, supra note 71, at 778.
Navy Investigative Service.
75
More recently, Australian intelligence was reportedly
frustrated by the degree to which British and American intelligence flooded it with
untested, often poor, intelligence in the run-up to the Iraq war, and considered
expanding its own collection abilities to reduce its dependence on other intelligence
agencies.
Consequently, other network actors can exercise a powerful form of
accountability. Unlike the public, intelligence agencies have the knowledge and
means to demand information and compliance from their intelligence sharing
partners.
76
Although the secrecy of transnational sharing makes it impossible to
evaluate the extent to which network sanctioning results in compliance with
professional norms, reputational control is not necessarily less effective than
hierarchical control.
77
When all network members subscribe to a norm, internal
regulation may be even more effective and efficient than external controls.
78

Nevertheless, as the next section will show, the lack of democratic oversight and
control of intelligence networks permits abuses and undermines human rights in a
way that calls out for democratic scrutiny and stringent regulation.
II. CHALLENGES OF TRANSNATIONAL INTELLIGENCE COOPERATION
FOR DEMOCRACY AND HUMAN RIGHTS
The self-regulating and secret nature of transnational intelligence networks
poses a significant challenge to our very conception of democracy and the
appropriate role of agencies in the democratic state. A democracy requires that the
people set policy. Thus, in liberal democracies, intelligence agencies have the sole
function of preserving the democratic state.
79
They are to carry out and inform
policy, not to make it. Democracy also entails transparency and oversight of
executive agencies and demands accountability, that is, clear standards to guide
agency behavior. The coercive power of intelligence agencies is, therefore, to be
reined in through statute, treaty, and legislative oversight.
Yet, as this Part argues, the involvement of our intelligence agencies in
information sharing networks challenges each of these assumptions. Section A
describes the lack of democratic accountability and oversight inherent in these
intelligence networks. The opacity with which these networks operate results in
75. Ephraim Kahana, Mossad-CIA Cooperation, 14 INTL J. OF INTELLIGENCE &
COUNTERINTELLIGENCE 409, 414 (2001); PAT M. HOLT, SECRET INTELLIGENCE AND PUBLIC POLICY: A
DILEMMA OF DEMOCRACY 127 (1995) (discussing the shake-up in CIA-Mossad relations after the arrest
of Jonathan Pollard for spying in the U.S. on behalf of Israel).
76. Robert O. Keohane, The Concept of Accountability in World Politics and the Use of Force, 24
MICH. J. INTL L. 1121, 112930 (2003) (Unlike outsiders, they can identify who is responsible for results
or for failures: they have information as a result of their organizational activity. Precisely because they
are likely to suffer if their organization does badly (insofar as accountability operates at the level of the
organization), they have incentives to help correct the problem at the individual level.).
77. Id. at 1134.
78. Koh, supra note 60, at 1401 (arguing that the most effective form of law-enforcement is not the
imposition of external sanction, but the inculcation of internal obedience). See also ORAN R. YOUNG,
INTERNATIONAL COOPERATION: BUILDING REGIMES FOR NATURAL REGIMES AND THE ENVIRONMENT
71 (1989) (Most members of social systems comply with the dictates of prevailing institutions most of the
time for reasons having little or nothing to do with their expectations regarding the imposition of
sanctions.).
79. PETER GILL & MARK PHYTHIAN, INTELLIGENCE IN AN INSECURE WORLD 14849 (2006).
public ignorance and legislative indifference, leaving services largely to determine
their own transnational activities even in liberal democracies. Instead of being
checked by legislation, information sharing is only minimally regulated by
professional standards. Section B argues that without proper oversight, intelligence
networks permit domestic foreign policy to be undermined and national and
international law to be circumvented. Paradoxically, as Section C explains, Western
intelligence agencies attempt to carry out their mission of preserving the democratic
state by cooperating with unethical intelligence agencies in illiberal states and
lending their support to authoritarian regimes. Moreover, the professional norms
agencies set for themselves in these networks prove easy to disregard. With
connections to less-reputable agencies, intelligence services from liberal democracies
often give in to the temptation to commit human rights abuses and to outsource
torture. Even the most professional agencies may trade in unreliable or untested
information, to the detriment of the demos.
A. Deficit of Democratic Accountability and Control
A functioning democracy assumes that each executive agencys actions are
subject to executive and legislative approval and control. Yet, intelligence networks
effectively shield ever-greater degrees of government activity from public view and
domestic structures of accountability.
80
With secrecy and non-disclosure as its
cardinal rules, transgovernmental intelligence cooperation often occurs without the
knowledge of domestic constituencies. The coercive power and substantial potential
for error of intelligence networks are subject to little or no domestic oversight or
review. As to terrorism or arms smuggling, the issue of accountability is especially
salient, because measures are usually taken against individuals (often citizens of
nations lacking strong accountability mechanisms) with little possibility of effective
political or legal challenges. Although networks exert some degree of accountability
through professional standards and reputational sanctioning, from a democratic
standpoint the view of intelligence as a profession that largely governs itself
according to its own definition of responsibility is alarming.
81

This section argues that the network connections of agencies of liberal
democracies currently operate in a legal void, an anathema to the rule of law and
democracy. As Section 1 explains, the secrecy and opacity with which these
networks function marginalize democratic institutions. Section 2 shows that
democratic oversight at any level is minimal, which in turn means intelligence
agencies have few clear statutory or regulatory limits on activities conducted with
their foreign counterparts.
80. Although scholars differ on the exact definition of accountability, Robert Keohane explains that
[a]ll satisfactory definitions of accountability include, explicitly or implicitly, two essential features:
information and sanctions. Keohane, supra note 76, at 1124. More specifically, there must be some
provision for interrogation and provision of information, and some means by which the accountability-
holder can impose costly sanctions on the power-wielder. Id.
81. Bruneau & Dombroski, supra note 31, at 166.
1. Lack of Transparency
At the heart of the lack of democratic accountability lies the fact that even
though intelligence agencies regularly cooperate with one another, their network
arrangements are nearly invisible to national publics, legislators, and international
bodies.
82
The opacity and deniability under which networks operate insulate each
intelligence agency from criticism. An agency can seldom be reprimanded for
sharing or failing to share information with network partners, since it denies having
them; nor does it risk major repercussions for other states intelligence failures
caused by bad intelligence it provided. Put simply, an oversight committee cannot
provide input, request information, or impose restraints on arrangements of which it
has no knowledge.
Cooperation can take place without public knowledge, legislative consent, or
even executive approval.
83
In 1954, for example, the Danish intelligence chief
approved and then covered up United States intelligence reconnaissance flights
across Danish territory, which could not be cleared through diplomatic channels.
84

Similarly, Turkish government leaders were not informed of the secret agreement
between Turkish and American military intelligence agencies to gather SIGINT.
85

More recently, the Lithuanian state security service gave permission to the CIA to
run a secret prison in the country, without informing the president or the prime
minister.
86

The perpetual secrecy of information shared through networks further
exacerbates the problem. Under current arrangements, any intelligence provided in
confidence by another intelligence agency stays perpetually secret and exempt from
freedom of information requests, unless the originator consents to its declassification
and release.
87
Within tightly knit intelligence agreements such as the UKUSA
arrangement, domestic and shared information may become commingled, resulting
in the more restrictive rule, intended for foreign information, to become the default
82. Aldrich, supra note 2, at 53 ([I]n a global era, when clandestine agencies rarely work alone on
large issues, the near invisibility of liaison arrangements to oversight by elected representatives is
problematic. Oversight mechanisms have not kept pace with global issues.).
83. Matthew M. Aid & Cees Wiebes, Conclusions, in SECRETS OF SIGNALS INTELLIGENCE DURING
THE COLD WAR AND BEYOND 313, 324 (Matthew M. Aid & Cees Wiebes eds., 2001) ([I]ntelligence
chiefs of many of the European Sigint organisations sometimes kept many of the details of their
intelligence collaboration with the US and Great Britain a secret from senior civilian officials in their own
governments.). At other times, political sensitivity created incentives to shield intelligence sharing from
the public eye. Id. at 326 ([W]hile the working relationship between NSA and the CIA on one hand and
the Norwegian military intelligence organisation, the FO/E, on the other hand was quite close, the political
leaders of Norway were constantly fearful that this relationship would become a matter of public
record.).
84. Alf R. Jacobsen, Scandinavia, Sigint and the Cold War, in SECRETS OF SIGNALS INTELLIGENCE
DURING THE COLD WAR AND BEYOND 209, 228 (Matthew M. Aid & Cees Wiebes eds., 2001).
85. Aid & Wiebes, supra note 83, at 324 (American military officials in Turkey had to remind their
colleagues back in Washington . . . not to mention [this agreement] in any discussions with Turkish
political or diplomatic authorities).
86. Milda Seputyte, Lithuania Let CIA Use Secret Prison for Interrogation, BLOOMBERG, Dec. 22,
2009.
87. See, e.g., Roberts, supra note 39, at 353357 (listing examples of the continuing influence of
NATOs security of information policy over domestic access-to-information laws accommodation of
originator control).
rule for classification and access to information.
88
Based on a study of NATOs
information sharing, Alasdair Roberts concluded that [t]he connection of
governmental networks means that the flow of information through any one
government will be increased, thus increasing incentives to tighten security and
classification of all information.
89
Thus, more information becomes permanently
withheld from those outside the intelligence community.
90
Even when the originator
or sharing agreement provides for declassification after a certain number of years,
declassification is not linked to the existence of a justification for continued secrecy.
91

2. Absence of Democratic Oversight
A constant throughout the globe is that intelligence agencies are considerably
insulated from sunshine and sanctions as compared to other government agencies.
Even in the United States, which is considered to have some of the most stringent
legislative monitoring of intelligence agencies,
92
Congress typically only reviews
intelligence policy after an issue becomes publicputting out the fire rather than
preventing it. Within Europe, some states do not have any parliamentary body to
monitor intelligence, although some employ other means of supervision.
93

88. See id. at 359. Roberts concludes that intelligence exchange agreements challenge assumptions
that global integration is favorable to increased transparency and that domestic policy on matters of
state secrecy is increasingly constrained by the thickening web of agreements on security of information.
Id. at 332.
89. Id. at 359 (Since the end of the Cold War, Canada is said to have developed more bilateral
intelligence relationships, and arguably, a more complex set of sensitivities regarding the protection of
information provided in confidence.). Romanian intelligence, for example, informed parliament that
NATO required that a bill for the classification and protection of shared intelligence as state secrets be
passed. Id. at 332 n.10; see also Duke, supra note 43, at 60910 (discussing increased classification and
very secret designations required of the EU to gain access to information from member states or other
organizations by NATO). Even international tribunals are forced to become more secretive. See
Chesterman, supra note 36, at 1122 (observing that the International Criminal Tribunal for the former
Yugoslavia, for example, created a national security exemption to its obligation to produce documents
and information).
90. WALTER LAQUEUR, THE USES AND LIMITS OF INTELLIGENCE 209 (1993) (stating that it is not
readily obvious why, many years after the event, intelligence files should remain closed. Some suspect
that this has less to do with the defense of the realm than with the wish to cover up past mistakes.);
Roberts, supra note 39, at 35354 (noting that international information sharing agreements cause the
Canadian government to continue to deny requests for access to information provided by other states even
where disclosure would not cause harm).
91. Under EU regulations, for example, the member state originator determines declassification;
originators may either state a date upon which a document may be declassified or review the classification
level every five years. Bjrn Mller-Wille, Improving the Democratic Accountability of EU Intelligence, 21
INTELLIGENCE & NATL SECURITY 100, 122 (2006). Nonetheless, because of lack of enforceability,
declassification rests fully on the will of the originator, most notably the Member States intelligence
communities. Id.
92. WILLIAM J. DAUGHERTY, EXECUTIVE SECRETS: COVERT ACTION AND THE PRESIDENCY 29
(2004) (stating that according to the former Director of Central Intelligence, [i]n no other country
including the parliamentary democracies of Western Europehas intelligence been subject to so much
investigation and review by the legislative branch as it has in the United States).
93. EUR. PARL., Comm. on Foreign Affairs, Human Rights, Common Security & Defence Policy,
Temp. Comm. on the ECHELON Interception System, European Parliament Resolution on the Existence
of a Global System for the Interception of Private and Commercial Communications (ECHELON
Interception System), para. K, INI/2001/2098 (May 9, 2001), available at http://www.cyber-
rights.org/interception/echelon/European_parliament_resolution.htm [hereinafter European Parliament
Central to the absence of democratic oversight and control is the fact that most
intelligence agencies, even in Western democratic states, operate either under
ambiguous statutes or without any statutory authorization. In Britain, the United
States, and Canada, at least one particular intelligence agency has operated (or
continues to operate) without any statutory authority or limitations.
94
This
represents a near total absence of accountability, the very concept of which
presuppose[s] norms of legitimacy that establish, not only the standards by which
the use of power can be judged, but also who is authorized to wield power and who is
properly entitled to call the power-wielders to account.
95
The absence of limits set
by a democratic body poses a serious danger given intelligence agencies coercive
power and capacity for deception.
96

Where legal restraints do exist, they impose limits almost exclusively on
domestic activity. These include requirements that foreign and domestic intelligence
be separated, residents and citizens information not be intercepted, and, as applies
to intelligence networks, residents and citizens data be shared only in accordance
with domestic data protections.
97
Even in democracies, however, there are generally
no statutory permissions or limitations on intelligence work outside national borders
or intelligence relations to foreign counterparts.
98
From a practical standpoint, this
indicates a lack of legislative involvement in setting the proper powers and limits of
intelligence. It also confounds democratic accountability because there are no clear
standards of what an agency is permitted to do, with whom it may form connections,
and under what circumstances intelligence sharing or other operations are
authorized.
Similarly, to the extent they do oversee intelligence agencies, national systems
by and large fail to address the transnational relationships and activities of their
intelligence agencies. Traditional democratic mechanisms used to attain information
and impose sanctionsnamely, oversight, legal/judicial constraints, and publicity
from the media, NGOs, and citizenspay little attention to transnational intelligence
cooperation.
99
For example, it was only after the European Parliament expressed
Resolution on ECHELON].
94. See TURNER, supra note 41, at 32 (noting that the existence of the National Reconnaissance
Office was only revealed in 1994); Stuart Farson, Canadas Long Road from Model Law to Effective
Political Oversight of Security and Intelligence, in WHOS WATCHING THE SPIES? ESTABLISHING
INTELLIGENCE SERVICE ACCOUNTABILITY 99, 101 (Hans Born et al. eds., 2005) (noting that the CSEs
existence was only acknowledged in the mid-1970s); BIRKINSHAW, supra note 69, at 50 (noting history of
lack of statutory authority for U.K. intelligence).
95. Grant & Keohane, supra note 59, at 30.
96. BIRKINSHAW, supra note 69, at 50 (It could be argued that, from a constitutional point of view,
one of the most glaring omissions is the absence of a precise statutory code covering the powers of the
services.); Mller-Wille, supra note 91, at 103 (It is crucial that laws regulating the services activities are
credible and adequate, i.e. that they allow agencies to do what is required to safeguard the democratic
system and the people. For the sake of checks and balances, these special powers must be regulated by
law and not simply by governmental decrees.).
97. As Stansfield Turner, former Director of Central Intelligence, observed, the laws and rules apply
mostly to interference with Americans and hence do not greatly affect most foreign intelligence espionage
operations. STANSFIELD TURNER, SECRECY AND DEMOCRACY: THE CIA IN TRANSITION 152 (1985).
98. See, e.g., Ian Leigh, Accountability of Security and Intelligence in the United Kingdom, in WHOS
WATCHING THE SPIES? ESTABLISHING INTELLIGENCE SERVICE ACCOUNTABILITY 79, 82 (Hans Born et
al. eds., 2005) (noting that legislation creating authority for U.K. intelligence failed to detail arrangements
for international cooperation).
99. Currently, we depend on the media, NGOs, and social networks to reveal abuses and activities of
our own intelligence agencies. Ronald J. Deibert, Deep Probe: The Evolution of Network Intelligence, 18
concern over the ECHELON system that the democratic bodies in the UKUSA
countries showed any interest in transnational intelligence agreements. The U.S.
investigation into the arrangement only began after illegal interceptions of citizens
communications were alleged.
100

Apathetic oversight bodies combined with few statutory restraints make
intelligence networks and their activities outside the domestic sphere the area of
weakest oversight and thus accountability.
101
As we will see, the failure to take
transnational relationships into account can render statutes limiting intelligence
activity in the domestic sphere practically toothless as well.
The transnational aspect of intelligence work can also stymie national efforts at
regulation and oversight. Oversight bodies are often specifically excluded from
receiving information exchanged through networksdepriving them of the potential
to review the use of network arrangements.
102
The Arar Commission, a Canadian
public inquiry into the extraordinary rendition of Syrian-born Canadian citizen
Maher Arar to Syria, is a case in point.
103
Without the cooperation of Canadas
intelligence partners, the findings were inevitably incomplete and the appropriate
officials and agencies could not be held accountable for wrongdoing. A Dutch
government inquiry into intelligence failures at Srebrenica faced similar obstacles to
uncovering the role of allied intelligence.
104
Likewise, although the U.S. House
Permanent Select Committee on Intelligence investigated, requested documents, and
held hearings on ECHELON, they were denied information due to its transnational
nature.
105

INTELLIGENCE & NATL SECURITY, 175, 175, 18687 ([M]any citizen networks have thrived on the
circulation of information concerning the signals intelligence capabilities and intelligence sharing
arrangements among the U.S., UK, Canada, Australia and New Zealand.); JOHNSON, supra note 35, at
211 (noting that were it not for a leak of the secret Iran arms sale operation in a Middle Eastern magazine
the Iran contra affair would not have been uncovered); LAURENCE LUSTGARTEN & IAN LEIGH, IN FROM
THE COLD: NATIONAL SECURITY AND PARLIAMENTARY DEMOCRACY xi (1994) ([J]ournalists are often
the only people able to gain access . . . . The media is also the only channel by which information from
insiders may reach the public.). Nevertheless, the media provide an inconsistent means of accountability.
Their inquiries remain principally domestic; they cannot impose the sanctions which are so important for
effective control of agencies. By necessity, they rely heavily on leakswhich, in the best-case scenario are
motivated by conscientious objections to wrongdoing, but in the worst, are selective or officially
sanctioned in order to support official policy. See Len Scott, Secret Intelligence, Covert Action and
Clandestine Diplomacy, 19 INTELLIGENCE & NATL SECURITY 322, 326 (2004) (stating that [t]he study of
intelligence requires consideration of the motives and agendas of sources and how far they can be
dissociated from the substance of what they provide). For example, in order to justify Maher Arars
continued detention, both Syrian Military Intelligence and the Canadian intelligence agencies leaked
inaccurate, misleading information to journalists. Ronald-Frans Melchers, The Maher Arar Case:
Implications for Canada-U.S. Law Enforcement Cooperation, 2006 J. INST. JUST. INTL STUD. 37, 3940
(2006). The leak was entirely self-serving and the information unfounded. Id.
100. Sloan, supra note 52, at 1487. In Canada, another UKUSA partner, Echelon is yet to reach
the parliamentary radar screen. Stuart Farson, Parliament and its Servants: Their Role in Scrutinizing
Canadian Intelligence, in AMERICAN-BRITISH-CANADIAN INTELLIGENCE RELATIONS 1939-2000 249
(David Stafford & Rhodri Jeffrey-Jones eds., 2000).
101. Mller-Wille, supra note 91, at 107.
102. BIRKINSHAW, supra note 69, at 4243 (noting the director of the GCHQ can refuse to disclose
information to the UK Parliamentary Committee on Security and Intelligence with oversight over GCHQ
because it was provided by, or by an agency of, the government of a territory outside the UK).
103. Maherarar.ca Home Page, http://www.maherarar.ca/ (last visited April 21, 2010).
104. Aldrich, supra note 2, at 53.
105. Sloan, supra note 52, at 1487.
Because secret written agreements, informal understandings, and professional
standards govern network exchanges of intelligence instead of legislation, the
judiciary is also almost completely excluded from regulating an intelligence agencys
foreign activities and contacts with foreign counterparts. Although domestic courts
rarely visit the issue of intelligence shared through networks, when legal issues
implicit in sharing arrangements have arisen, international obligations to network
partners have typically trumped judicial scrutiny and the rights of criminal
defendants.
106
For instance, in the U.K. trial of the IRAs Nicholas Mullen,
intelligence officers withheld information about his allegedly illegal deportation by
Zimbabwean intelligence agents from the court and government officials.
107
Most
recently, however, in a decision that may indicate courts reassertion of their role in
overseeing intelligence, the British Court of Appeals rejected the British
governments argument that release of information regarding the torture of a former
Guantanamo prisoner should be kept secret out of concern for its intelligence
sharing relationship with the CIA.
108
As a general matter, however, the result of
information sharing arrangements is to deny domestic actors, including the courts,
the opportunity to make their own decisions about the disclosure of information
within a certain policy domain.
109

B. Collusion to Avoid Democratically Determined Policies and Statutes
In practice, a unique danger of intelligence networks lies in their tendency to
encourage intelligence agencies to collaborate with one another to the detriment of
the interests of the democratic nations they are meant to serve. Because of the
corporatism agencies exhibit through networks, the closeness of the practitioners to
each other may be greater than to the precise policy objectives and interests of the
organizations or states to which they formally belong.
110
This is especially likely to
occur among tight-knit intelligence services. In the UKUSA intelligence community,
elements of [the community] frequently come to perceive their ultimate loyalties as
lying more with the UKUSA community than with their own governments.
111
For
example, Australian army officers working with the CIA during the Vietnam War
were prepared to swear not to divulge details about their joint activities to their own
commanding officers and government.
112
The closeness of ties has sometimes been
used counter to the policies of the agencies own governments.
113
For instance,
106. Martin Rudner, Canadas Communications Security Establishment from Cold War to
Globalization, in SECRETS OF SIGNALS INTELLIGENCE DURING THE COLD WAR AND BEYOND 97, 12324
(Matthew M. Aid & Cees Wiebes eds., 2001) [hereinafter Rudner, Canadas Communications Security
Establishment] (The legal issues implicit in Sigint-derived evidence have never been tested before
Canadas courts. Whenever questions have been raised, mere reference to Canadas international
obligations has sufficed to defer detailed inquiries.).
107. DORRIL, supra note 71, at 768.
108. Scott Horton, British Appeals Court Forces Release of Torture Details, HARPERS, Feb. 11, 2010,
available at http://www.harpers.org/archive/2010/02/hbc-90006521.
109. Roberts, supra note 39, at 355.
110. GILL, supra note 2, at 37. See also Philip B. Heymann, International Cooperation in Dealing with
Terrorism: A Review of Law and Recent Practice, 6 AM. U. J. INTL L. & POLY 1, 2 (1990) ([I]nstitutional
constituencies such as law enforcement officials may find that their interests resemble those of their
foreign counterparts more than those of other groups within their own country.).
111. RICHELSON & BALL, supra note 12, at 305.
112. Id.
113. Id.
intelligence professionals in various countries were aware of or suspected each
others knowledge of weapons-smuggling rings operating out of Pakistan to supply
weapons to countries including Libya and North Korea, yet they kept this
information from the populations that they purported to protect.
114

Network partners have also colluded to avoid statutory restraints on domestic
activity. Statutes prohibiting eavesdropping on citizens or residents without a
warrant, or creating a barrier between intelligence and law enforcement, have been
circumvented. The German BND, for instance, reportedly uses the European
Counter-Terrorist Intelligence Center in Paris (where agents from Britain, France,
Germany, Canada, Australia, and the United States work together against terrorism)
to read information from German law enforcement agencies it would be barred from
obtaining at home as a matter of domestic law.
115
In Norway, the CIA was given
permission to penetrate Muslim groups and investigate individuals on Norwegian soil
without normal review and accountability procedures in place and with little, if any,
control on the part of Norwegian authorities.
116

Numerous revelations involving the UKUSA arrangement disclose the potential
for circumventing statutory and constitutional law. Within ECHELON, the partner
agencies all submit a list of keywords; every partner station in the world then collects
telephone, fax, email, and other electronic communications (and, it is suspected,
internet traffic) in which the keywords appear and then sends them to the requesting
agency.
117
Likely as part of this program, at least three of the major telephone lines in
Great Britain, each capable of carrying 100,000 calls, are wired through the NSA
listening station, allowing direct taps into British Telecoms network.
118
The result is
that the collecting agency does not necessarily know what its stations are collecting.
119

Reports suggest that UKUSA agencies also purposely use ECHELON to
exchange surveillance information on each others citizens in violation of domestic
statutes.
120
NSA employees have provided details of the use of the network to
114. RON SUSKIND, THE ONE PERCENT DOCTRINE: DEEP INSIDE AMERICAS PURSUIT OF ITS
ENEMIES SINCE 9/11 269 (2006) (In the interplay between decisions made by the intelligence
professionals and a small circle of policy makersand the competing claims made by other branches of
government or by the public, with its recognized right to understand what truly guides U.S. foreign
policyalmost all the options reside with the parties of the first part.).
115. Dana Priest, Help from France Key in Covert Operations: Pariss Alliance Base Targets
Terrorists, WASH. POST, July 3, 2005, A1; European Parliament Resolution on ECHELON, supra note 95,
para. J ([W]hereas the Member States cannot circumvent the requirements imposed on them by the
ECHR by allowing other countries intelligence services, which are subject to less stringent legal
provisions, to work on their territory, since otherwise the principle of legality, with its twin components of
accessibility and foreseeability, would become a dead letter . . . .). Privacy protections also risk being
violated. See generally Francesca Bignami, Towards a Right to Privacy in Transnational Intelligence
Networks, 28 MICH. J. INTL L. 663 (2007) (stating that privacy is one of the most critical liberal rights to
come under pressure from transnational intelligence gathering).
116. Rudner, Hunters and Gatherers, supra note 10, at 215.
117. HAGER, supra note 52, at 29; Sloan, supra note 52, at 147678 (listing various types of methods
and communications intercepted through the ECHELON system); Rudner, Britain Betwixt and Between,
supra note 12, at 58182.
118. Lawner, supra note 11, at 453.
119. HAGER, supra note 52, at 29 (This means that the New Zealand stations are used by the
overseas agencies for their automatic collectingwhile New Zealand does not even know what is being
intercepted from the New Zealand sites for the allies.).
120. Seth F. Kreimer, Watching the Watchers: Surveillance, Transparency, and Political Freedom in
intercept politicians telephone calls.
121
CSE also monitored communications of U.K.
Prime Minister Margaret Thatchers cabinet members on behalf of the GCHQ.
122

Several GCHQ officials similarly revealed that partners were targeting and
distributing information regarding peaceful political groups.
123
More recently, at a
meeting of the UKUSA intelligence heads, U.S. Director of Central Intelligence
George Tenet described statutes against eavesdropping as among the shackles that
would, at the very least, be loosened, if not in practice discarded.
124
He suggested, as
Suskind describes, that [a] country may not be able to tap the lines of its own
citizens without legal authorization. But theres nothing to stop it from listening in
on some other countrys citizen, and then filing very thorough reports to that foreign
citizens government. Just as long as the report does not hand over the specific raw
matterthe SIGINT dispatch of nouns and verbsthe letter of various privacy laws
would stay intact.
125

Where there is peer accountability, rather than outside oversight, as these
examples demonstrate, network norms will check the power of network partners
only insofar as abuses are against the interests or principles of the other entities
within the transgovernmental networks.
126
Consequently, there is a danger that
congruence of goals and close cooperation within intelligence networks will lead to
collusion against the interests of outsiders rather than improved professional
norms.
127
The next section will examine several such instances.
C. Undermining Foreign Policy and Human Rights through Connections to Illiberal
Agencies
Just as intelligence cooperation imperils legal safeguards, so too does it run the
risk of intelligence services working with each other against the interests of their
domestic publics. In particular, cooperation with intelligence agencies in regimes
that have little in common with liberal democracies threatens the promotion of
democracy and human rights. As Section 1 argues, information exchange with
authoritarian regimes comes with a price. A quid pro quo is expected (and has been
provided) by professionalized intelligence services and democratic nations. As
Section 2 demonstrates, less-reputable agencies have profited from their inclusion in
the War on Terror, 7 U. PA. J. CONST. L. 133, n.114 (2004).
121. See, e.g., Sloan, supra note 52, at 148586 (discussing the claim by an NSA contract employee to
have witnessed firsthand the real-time interception of a telephone call made by United States Senator
Strom Thurmond).
122. Rudner, Canadas Communications Security Establishment: From Cold War to Globalisation 15
(Ottawa: Norman Paterson Sch. of Intl Affairs, Carleton Univ., Occasional Paper No. 22, 2000) (noting
that due to the political sensitivity, GCHQ could not be directly involved, so CSE intercepted the
communications at the Canadian High Commission in London and delivered the information to the
GCHQ).
123. Patrick S. Poole, CENTER FOR TECHNOLOGY POLICY, ECHELON: AMERICAS SECRET
GLOBAL SURVEILLANCE NETWORK (1999/2000), available at http://hp.kairaven.de/miniwahr
/freecongress1.html (describing the technical workings of ECHELON in detail).
124. SUSKIND, supra note 114, at 85.
125. Id.
126. Grant & Keohane, supra note 59, at 39.
127. Id.; Herman, supra note 3, at 351 (confirming that since 9/11, the balance has tipped against
ethical restraints, and commenting that [i]f the wartime metaphor fits counter-terrorism, it implies
relatively few moral restrictions on information gathering on its targets.).
intelligence sharing networks to crack down on dissenters, demand concessions from
the West, and continue systematic torture and indefinite detention. While admitting
that these partners engage in rights violations, numerous commentators suggest that
no one is harmed by the exchange of information.
128
Section 3 aims to challenge this
assumption. The problem is not simply illiberal regimes engaging in violations of
individual rights, but also liberal democracies encouraging and soliciting violations
for which they are not held accountable.
1. Quid Pro Quo: Intelligence at What Price?
Partner services in illiberal regimes do not offer intelligence out of altruistic
motives; like any other intelligence partner, they want something in return. Most
commonly, less reputable agencies demand repayment in kind. Russian intelligence
warned the United States that information sharing could not be one-way traffic.
129

The desirability of sharing with authoritarian regimes, however, depends on the type
of information requested. It is one thing for Canadian intelligence to supply an
assessment on the potential for terrorists to use the avian flu virus as a biological
weapon with the Libyan, Saudi Arabian and Egyptian intelligence services.
130
It is
quite another to provide intelligence on individual Chechens to Russian intelligence.
Illiberal agencies may also demand that intelligence services in liberal
democracies spy on migr or dissident groups on their behalf. South Africa, for
example, was tipped off to the activities of the African National Congress and the
location of Nelson Mandela by intelligence services in Western democracies whose
publics opposed apartheid.
131
In return for providing the CIA with information on
Libyan nationals with ties to international terrorists, Libyan intelligence was allowed
to interrogate prisoners at Guantanamo Bay about exiles in London.
132
The FBI also
arrested and interrogated one of Qadaffis primary opponents and long-time critic of
al Qaeda based on information provided by Libya.
133
Unsavory agencies have come
to expect such cooperation; British intelligence faced complaints from Egyptian and
Jordanian intelligence for failing to act on their requests for information on migr
communities.
134

128. Peter Gill, Securing the Globe: Intelligence and the Post-9/11 Shift from Liddism to Drainism,
19 INTELLIGENCE & NATL SECURITY 467, 477 (2004) (Transnational information exchange is one thing,
brokering the use of torture is another.); Herman, supra note 3, at 342 (observing that some say that
[i]ntelligence is information and information gathering, not doing things to people; no-one gets hurt by it,
at least not directly).
129. Julian Borger & Richard Norton-Taylor, Dirty War that Could Prove Decisive, GUARDIAN
(U.K.), Oct. 2, 2001, at 2.
130. Arar Commission ReportA New Review Mechanism, supra note 4, at 142.
131. HOLT, supra note 75, at 72 (contrasting this intelligence exchange with intelligence on Soviet
submarine and other ship movements around the Cape of Good Hope in exchange for information on
Soviet and Cuban activities in Angola). Cooperation with apartheid South Africa posed a dilemma for
Western intelligence, whose governments officially opposed apartheid. MICHAEL HERMAN,
INTELLIGENCE SERVICES IN THE INFORMATION AGE: THEORY AND PRACTICE 40 (2001).
132. Ken Silverstein, How Kadafi Went from Foe to Ally, L.A. TIMES, Sept 4, 2005, available at
http://articles.latimes.com/2005/sep/04/world/fg-uslibya4/1.
133. Id. at 10.
134. Rudner, Hunters and Gatherers, supra note 10, at 214. Jamie Wilson et al., New Brothers in
Armsand Cash and Intelligence, GUARDIAN (U.K.), Oct. 20, 2001, at 7 (Britains contribution is
expected to include the granting of Russian demands that a hard line be taken against Chechen exiles in
A partner agency may also expect Western intelligence to turn a blind eye to its
activities in their territory. For instance, out of fear of jeopardizing relations with
Iranian intelligence under the Shah, the U.S. intelligence community tolerated
operations against Iranian dissidents in the United States.
135
Iran later used the
information gained from these operations to take reprisals against dissidents families
in Iran.
136
Chinese intelligence cooperation has reportedly generated German
intelligence support for surveillance of the Chinese democratic opposition abroad.
137

Through these arrangements, intelligence services in liberal democracies help silence
domestic and international dissent against authoritarian governments.
138

Another preferred method of compensating network partners is to provide
them with funds and equipment. This may seem ethically neutral but often serves to
prop up authoritarian regimes and subvert democracy. For example, equipment
provided by the German BND was used by: the Indonesian military intelligence
service to overthrow President Sukarno; Ugandan dictator Idi Amin to eavesdrop on
opponents; a guerrilla organization to destabilize the legal Mozambique government;
and, more recently, Chinese intelligence to monitor the movements of journalists and
dissidents.
139
Similarly, British intelligence officials report that, in return for
information on terrorist groups, states like Malaysia receive surveillance equipment
they can use against dissidents in their own territory.
140
Just a month after 9/11,
intelligence partners, like Egypt and Oman, received arms from the United States for
their cooperation.
141
The United States has also built counterterrorist intelligence
centers for numerous foreign agencies, hoping to gain valuable intelligence from
agencies the CIAs former Deputy Director for Operations described as utterly
unhesitant in what they will do to get captives to talk.
142

2. Silencing Criticism and Bolstering Authoritarian Regimes
When their intelligence agents aid antidemocratic and human-rights abusive
practices of friendly dictators in hopes of good intelligence, liberal democracies
sacrifice their leverage to promote rights and damage their reputations at significant
political cost.
143

Criticism of repressive governments with cooperative intelligence agencies has
dramatically declined. In the past the U.S. condemned Malaysias detention of
dissidents under the Internal Security Act; today Malaysia is hailed as a beacon of
stability and the detentions have been praised.
144
Other liberal democracies have
similarly reduced public criticism.
145

London . . . .).
135. HOLT, supra note 75, at 129.
136. Id.
137. Schmidt-Eenboom, supra note 74, at 158.
138. HOLT, supra note 75, at 127.
139. Schmidt-Eenboom, supra note 74, at 15459 (listing more examples).
140. Wilson et al., supra note 134, at 7.
141. Id.
143. Harold Hongju Koh, The Spirit of the Laws, 43 HARV. INTL L.J. 23, 29 (2002).
144. INTL CAMPAIGN AGAINST MASS SURVEILLANCE, THE EMERGENCE OF A GLOBAL
INFRASTRUCTURE FOR MASS REGISTRATION AND SURVEILLANCE 46 (2005), available at
www.statewatch.org/news/2005/apr/icams-report.pdf [hereinafter INTL CAMPAIGN AGAINST MASS
As a result, in many countries, progress in improving human rights practices,
especially those of security services, has been reversed. For example, in Morocco, a
key U.S. intelligence ally, security services resumed indefinite detention of suspects
in secret interrogation centers, and the government enacted a broad anti-terrorism
bill so in which covers almost any violent crime.
146
By labeling dissidents or rebels as
terrorists, authoritarian states now legitimize repressive practices and shield these
practices from criticism by their partners.
147

Counterterrorism intelligence sharing with the West actively bolsters the power
of authoritarian dictators and their ruthless intelligence services. Due to the
cooperation of the brutal Uzbek security services, U.S. policymakers refer to Uzbek
President Islam Karimov as a partner and ally, strengthening his hand.
148
Indeed, one
of the reasons that we may now know so much more about the illiberal agencies
involved in information sharing is that certain foreign services disclose their ties to
Western intelligence agencies as a way of demonstrating their strength and shoring
up their position domestically.
149
Two instances of it was Libyas Qadaffi and Syrias
Assad, for example, who revealed their agencies intelligence sharing with the
West.
150

SURVEILLANCE]. The relationship between less criticism and greater intelligence cooperation may not be
apparent. However, Malaysias most substantial contributions to U.S. counterterrorism efforts seem to be
in the intelligence field; it has shared intelligence, offered access to detainees, and now hosts a Southeast
Asia center for terrorism training. Tay, supra note 18, at 119 (noting that Malaysia supplied intelligence
and detained numerous suspects on behalf of the U.S. and criticism about human rights subsequently
waned). Moreover, the presence of intelligence gathering facilities or reliable intelligence sources has
historically discouraged criticism of human rights abuses. HOLT, supra note 75, at 127.
145. INTL CAMPAIGN AGAINST MASS SURVEILLANCE, supra note 144, at 46 (Weeks after the 9/11
attacks, . . . democratic leaders like German Chancellor Gerhard Schroeder and Italys Prime Minister
Silvio Berlusconi were saying that they would have to judge Russian operations in Chechnya differently.);
AMNESTY INTERNATIONAL, HUMAN RIGHTS DISSOLVING AT THE BORDERS? COUNTER-TERRORISM
AND EU CRIMINAL LAW 2 (2005), available at http://web.amnesty.org/library/pdf/IOR610132005
ENGLISH/$File/IOR6101305.pdf [hereinafter AMNESTY INTERNATIONAL] (As far as cooperation with
third countries in the fight against terrorism is concerned, the EU and its Member States too often are
prepared to remain silent on the issue of rights protection.).
146. CAROL MIGDALOVITZ, CONG. RESEARCH SERV., RS 21579, MOROCCO: CURRENT ISSUES
(2005), available at http://www.policyarchive.org/handle/10207/bitstreams/3768.pdf; JOSEPH L.
DERDZINSKI, INTERNAL SECURITY SERVICES IN LIBERALIZING STATES TRANSITIONS, TURMOIL, AND
(IN)SECURITY 7273 (2009).
147. INTL CAMPAIGN AGAINST MASS SURVEILLANCE, supra note 144, at 46 (noting that on the
extension of Egypts emergency law in February 2003, the U.S. State Department stated that the U.S.
under[stood] and appreciate[d] the Egyptian governments commitment to combat terrorism and
maintain stability); Tay, supra note 18, at 120 (noting that the Indonesians government relabeling as
counterterterrorism its renewed military efforts in Aceh has received muted criticism from the U.S.).
Syrias 1982 massacre of 10,000 residents and the disappearance of thousands more in reaction to a Muslim
Brotherhood insurgency has been repackaged as a textbook antiterrorism campaign. Neil
MacFarquhar, Syria Repackages of Muslim Militants as Antiterror Lesson, N.Y. TIMES, Jan. 14, 2002, at
A8.
148. Thomas Carothers, Promoting Democracy and Fighting Terror, 82 FOREIGN AFFAIRS 84, 86
(2003).
149. HOLT, supra note 75, at 71. The practice of disclosing ties seems to have increased after
September 11.
150. Gawdat Bahgat, Transatlantic Cooperation: Libyas Diplomatic Transformation, 29 FLETCHER
F. WORLD AFF. 43, 47 (2005); Rudner, Hunters and Gatherers, supra note 10, at 217. Sudanese
intelligence officers brag that American intelligence considers us to be a friend and [t]he information
we have provided has been very useful to the United States. Silverstein, supra note 132, at A1.
Political favors that fly in the face of human rights policies of liberal
democracies have also been doled out. In the past, the U.S. rejected Sudans
overtures, refusing to cooperate with its intelligence service in fear it would
legitimize the repressive government.
151
Yet, days after CIA officials and Sudans
deputy intelligence chief concluded a secret intelligence sharing arrangement, the
U.S. abstained on a vote at the United Nations Security Council, freeing Sudan from
international sanctions.
152
Due to ties with Western intelligence services, Pakistan
also succeeded in lifting U.S. prohibitions on arms sales.
153

3. Complicity of Liberal Democracies in Rights Violations
Liberal states and their authoritarian partners are subject to various
international and domestic legal obligations that constrain the scope of action against
individuals. Citizens in many liberal democracies see themselves as devoted to
human rights and have enshrined protections against torture and other fundamental
rights violations into national law. Although it is expected that intelligence agencies
violate the domestic laws of other states in their collection activities, they are not
authorized to circumvent these international and domestic restraints.
154

Yet, as has become apparent since 9/11, intelligence agencies from liberal
democracies sometimes intentionally solicit torture or coercive action from illiberal
partners. At others, they innocently ask for help within the normal course of
investigations, but may be reckless as to the consequences of their requests. Always,
however, connections to illiberal regimes will raise the specter of complicity in or
encouragement of the violation of individual rights.
In their hunt for terrorist operatives, Western intelligence has depended on
counterparts in countries around the world to conduct interrogations and report
information gleaned.
155
Western intelligence agencies provide lists of questions and
151. Silverstein, supra note 132, at A1.
152. Id.
153. Wilson et al., supra note 134, at 7.
154. E.U. NETWORK OF INDEPENDENT EXPERTS IN FUNDAMENTAL RIGHTS, THE BALANCE
BETWEEN FREEDOM AND SECURITY IN THE RESPONSE BY THE EUROPEAN UNION AND ITS MEMBER
STATES TO THE TERRORIST THREATS 20 (Mar. 31, 2003), available at http://www.humanrights-
observatory.net/dhangles/tematic2.pdf (It is important to keep in mind that the Member States of the
European Union, which are party to the European Convention of Human Rights, are bound to comply
with that instrument, including in the context of interstate cooperation that they choose to enact with
nonmember States.). As a matter of international law, a state that solicits or assists in a human rights
violation is responsible for it. Report to the International Law Commission to the General Assembly, 56
U.N. GAOR Supp (No. 10) at 64, U.N. Doc. A/56/10 (2001), reprinted in [2001] 2 Y.B. Intl L. Commn 31,
U.N. Doc. AA/CN.4/SER.A/2001/Add.1 (Part 2). ([I]nternationally wrongful conduct often results from
the collaboration of several States rather than of one State acting alone.). See also id. at 66 (An aiding
or assisting State may not deliberately procure the breach by another State of an obligation by which both
States are bound; a State cannot do by another what it cannot do by itself.).
155. See HUMAN RIGHTS WATCH, REPORT TO THE CAN. COMMM OF INQUIRY INTO THE ACTIONS
OF CANADIAN OFFICIALS IN RELATION TO MAHER ARAR 7 (May 17, 2005) [hereinafter HUMAN RIGHTS
WATCH] (If we are getting everything we need from the host government, then theres no need for us to
[conduct interrogations], a former U.S. government official told Human Rights Watch. There are some
situations in which the host government can be more effective at getting information.). See, e.g., INTL
CAMPAIGN AGAINST MASS SURVEILLANCE, supra note 144, at 47 (noting the CIA station chief in
Tashkent readily acknowledged torture was deployed [in Uzbekistan] in obtaining intelligence [from U.S.
suspects]).
accept answers that are almost certainly coerced from the suspects, while knowing
that many of these agencies engage in acts of torture. Involvement is not limited to
U.S. intelligence services, but rather includes a wide variety of intelligence agencies
from democratic states.
156

In the most egregious cases, Western services have exploited the lack of
accountability and rights protection in partner nations to circumvent human rights
law and violate individual rights. The case of Maher Arar provides a well-known
example. Arar is a Canadian citizen entirely innocent of wrongdoing who was
detained at a U.S. airport, sent to Syria, and interrogated and tortured by Syrian
Military Intelligence (SMI).
157
Despite knowing that the SMI, an agency notorious
for its torture practices, had custody of a Canadian citizen, the Royal Canadian
Mounted Police (RCMP) offered large volumes of highly sensitive documents and
information, seized during investigative efforts or obtained from confidential
informants associated to terrorist cells operating in Canada.
158
The RCMP also sent
questions for Arar and other detainees in Syria against advice that the SMI would
likely use torture to seek answers and would consider the questions proof that the
individuals were in fact terrorists.
159
Arar was indeed torturedbeaten and kept
alone in an unlit cell the size of a grave.
160
A proper appreciation of the SMIs
practices and lack of professionalism should have suggested that answers obtained
from these interrogations would be untrustworthy. The SMI, for example, found
Arars salary and job description unbelievable.
161

In another instance, the U.K. security service MI5 solicited abuse from its
network partners in Africa. When several of acquaintances of cleric Abu Qatada
traveled to the Gambia, MI5 cabled a foreign intelligence agency, labeling them
Islamic extremists and disclosing their destination.
162
Upon arrival, CIA and
Gambian intelligence agents were waiting. When one of them, a British citizen,
156. See, e.g., Rudner, Hunters and Gatherers, supra note 10, at 219 (observing that Australia and
Southeast Asian partners submitted lists of questions to be put to terrorist suspect Hambali). Canadian
intelligence has also implicated individuals and sent questions to foreign counterparts. For example,
Maher Arars brother-in-law was questioned by the Tunisian police after moving there from Canada,
based on information to which only Canadian agents would have had access. In a similar incident,
Kassim Mohamed, who divides his time between Toronto and Egypt, was questioned by CSIS in Canada
after videotaping Toronto landmarks . . . . When he arrived in Egypt, he was arrested and held for two
weeks, handcuffed and blindfolded, in a prison in Cairo. INTL CAMPAIGN AGAINST MASS
SURVEILLANCE, supra note 144, at 26. See also Philip B. Heymann, Civil Liberties and Human Rights in
the Aftermath of September 11, 25 HARV. J.L. & PUB. POLY 441, 455 (2001) (noting that many abuses
occur in cases in which the CIA will know of the capture of individuals and may make known to the
international security apparatus of a foreign nation exactly what it would like to know).
157. See generally Arar Commission ReportAnalysis and Recommendations, supra note 46, at 32
(includes detailed account of Mr. Arars experience).
158. Id. at 38
159. Id.
160. Id. at 5657.
161. Id. at 56.
162. Craig Whitlock, Courted as Spies, Held as Combatants: British Residents Enlisted by MI5 After
Sept. 11 Languish at Guantanamo, WASH. POST, Apr. 2, 2006, at A1. See also HUMAN RIGHTS WATCH,
CRUEL BRITANNIA: BRITISH COMPLICITY IN THE TORTURE AND ILL-TREATMENT OF TERROR SUSPECTS
IN PAKISTAN 17 (2009), available at http://www.hrw.org/en/reports/2009/11/24/cruel-britannia-0
(investigating U.K. complicity in torture in Pakistan and concluding that it is inconceivable that security
services were not aware their partners were engaging in torture of detainees about whom they were
providing questions and receiving information).
asked to meet with the British consul, an American agent replied, Who do you
think asked us to arrest you? Where do you think this information came from, the
questions we are asking you?
163
All were detained for over a month. The two non-
British citizens were subsequently sent to Guantanamo Bay and remain there (MI5
has paid them several visits).
164
Because it only shared information, the United
Kingdom denies responsibility because it did not specifically request that the
individuals be detained.
165

Even where a Western agency attempts to mitigate rather than capitalize on the
tactics of disreputable or brutal agencies, the possibility of torture, disappearance, or
detention remains a serious risk. The tendency of intelligence officers to focus on
worst case scenarios increases the likelihood that information, will be acted upon
often to the detriment of individual human rights.
166
Repressive countries
intelligence services are particularly unlikely to exercise restraint when called upon
to conduct interrogations on behalf of Western services, especially where allegations
of Islamic terrorism are involved.
167
The simple act of providing questions about an
individual to a less professional agency can cause it to conclude that the sharing
agency considers the individual to be a serious terrorist threat.
168

The brutal nature of some partners of Western intelligence agencies blurs the
line between acceptable and unacceptable requests. For example, in an attempt to
confirm reports of the death of al Qaeda leader Zawahiri, a CIA operations manager
called an Egyptian intelligence chief for a DNA sample from Zawahiris brother,
who had been detained in Cairo.
169
The Egyptian officer reportedly responded, No
problem. Well get his brother, cut off his arm, and send it over.
170
Even well-
intentioned and necessary requests for intelligence may, therefore, create a ripple
effect beyond the requesting countrys borders with consequences that cannot be
controlled by that country.
171
Western intelligence should thus carefully evaluate
the risks posed by requesting information from illiberal services, meticulously phrase
requests for information, and attempt to ensure treatment of persons in accordance
with legal obligations. Unless they do so, requests for information may become, in
fact, requests for violation of human rights law.
172

163. Whitlock, supra note 162, at A1.
164. Id.
165. Id. Even if the United Kingdom did not specifically ask for the detention (which seems
unlikely), Western agencies provide specific intelligence with the intent that their partners act on it.
RONALD KESSLER, THE CIA AT WAR: INSIDE THE SECRET CAMPAIGN AGAINST TERROR 273 (2003)
(By the middle of 2002 the CIA had rolled up three thousand terrorists in a hundred countries. Usually a
foreign service made the arrest based on CIA information.).
166. HOLT, supra note 75, at 86. As a general matter, there are fewer repercussions for acting against
individuals or on terrorist information since it generally does not engage geopolitical concerns.
167. See Heymann, supra note 156, at 45354. ([P]rotections [available in the United States] are
often not available in anything like the same measure in states where terrorists are likely to seek haven.
Those countries internal structure and police apparati are likely to be far less constrained if activated by
the CIA on behalf of America.).
168. Arar Commission ReportAnalysis and Recommendations, supra note 46, at 39.
170. Id.
171. See National Security Background Paper to Arar Commission, supra note 26, at 16 (discussing
effect from Canadian perspective).
172. Heymann, supra note 156, at 455.
Western agencies also have subverted human rights and endorsed abuse by
receiving information from their new allies. The case of Craig Murray, former
British ambassador to Uzbekistan, is one example of Western knowledge of abuse.
When he urged the British Foreign Office to stop using intelligence elicited from
terrorism suspects through torture and other coercive means, he was told that the
intelligence could still permissibly be used even if it was elicited by torture, as long
as the mistreatment was not at the hands of British interrogators.
173

Using information obtained through torture or indefinite detention is not in the
interest of accuracy or effectiveness. Mistreatment typically results in information
replete with false positives,
174
on the basis of which innocent persons are then
detained and mistreated, continuing the cycle. The cost of receiving information of
dubious reliability from illiberal states is paid not only by persons in or sent to
repressive countries but also by residents of liberal democracies throughout the
world. As the Arar Commission determined, [r]eceipt of . . . information may lead
to significant personal consequences for individuals . . . such as surveillance, further
collection of personal information, or interrogation.
175

Services in repressive regimes are of questionable reliability because of their use
of torture. They are doubly dubious as their chief functions are to lend blind
allegiance to the regime and to repress dissidents.
176
They also often have
contradictory interests: providing financing and support to some designated terrorist
groups, and offering to work against others (or sometimes the same group).
Western services apparent credulity has already resulted in numerous errors.
A human source identified by the Libyan intelligence head was responsible for
allegations that Saddam Hussein provided biological and chemical weapons training
to al Qaeda.
177
The Bush administration acted upon it even though the CIA already
doubted the sources veracity.
178
New intelligence sharing arrangements between
U.S. intelligence and the Russian security intelligence services (FSB) have also
generated unreliable information and rights violations. A naturalized U.S. citizen,
173. Don Van Natta, Jr., U.S. Recruits a Rough Ally to be a Jailer, N.Y. TIMES, May 1, 2005, at A22.
174. Tony Pfaff, U.S. Army, Bungee Jumping Off the Moral Highground: The Ethics of Espionage in
the Modern Age, http://www.usafa.edu/isme/JSCOPE02/Pfaff02.html (It is a well-established fact that
information gained under the duress of torture is rarely reliable.). See also Shane OMara, Torturing the
Brain: On the Folk Psychology and Folk Neurobiology Motivating Enhanced and Coercive Interrogation
Techniques, 13 TRENDS IN COGNITIVE SCI. 497, 49798 (2009) (showing with scientific evidence that
coercive intelligence techniques and torture make it less likely for the subject to accurately recall
information, and more likely for false memories to replace real ones).
175. National Security Background Paper to Arar Commission, supra note 26, at 17 ([I]nformation
obtained from other countries may not have been acquired in ways consistent with rights and freedoms
protected [in Canada]. It may, for example, have been obtained through torture or other unacceptable
investigation techniques, or in the absence of checks and balances to ensure reliability.).
176. See Herman, supra note 5, at 229. Moroccan intelligence, for example, spends a great deal of its
resources on extensive surveillance of three particular classes of foreigners: US citizens (for their own
safety,) Spaniards, and journalists, the latter two to ensure they were engaging in approved activities.
DERDZINSKI, supra note 146, at 5859.
177. Peter Finn, Detainee Who Gave False Iraq Data Dies in Prison in Libya, WASH. POST, May 12,
2009.
178. See SUSKIND, supra note 114, at 18788 (discussing the CIAs concerns over intelligence on
WMDs in Iraq). This is not a new experience for the CIA and Mossad, which together created the Iranian
intelligence agency and the Korean Central Intelligence Agency (known for their heavy-handed brutality
and torture) and received much intelligence of dubious validity from both. HOLT, supra note 75, at 71.
Omar Shishani was accused of having links to terrorism, based on information
provided by the FSB.
179
The FSB and witnesses it supplied identified Shishani as an
Islamic radical who introduced Wahhabism to Chechnya in the 1990s. In the end,
this information was revealed to be inaccuratea case of false identity.
180
Yet, U.S.
intelligence appeared willing to accept this information unconditionally.
In the end, it is clear that the sort of compliance enforced through professional
network sanctioning does not necessarily produce strong compliance or good
behavior on the part of intelligence services. A real risk is the proliferation of lower,
rather than higher, standards. Because acculturation relies on the pull of peer
pressure rather than the legitimacy of the norm, it is value-neutral as to the norms
transmitted through the network. Thus, actors systematically conform (under the
right conditions) even if the group is clearly wrong . . . .
181
In the case of
transnational intelligence sharing networks, the norms have both been positive
good investigative tacticsand negativemost recently, torture and indefinite
detention.
182
Though professionalized intelligence services may generally be relied
upon to achieve high standards, greater cooperation with intelligence agencies of
dubious virtue has strained the professional ethos.
183

III. A PROPOSAL FOR MORE PROFESSIONAL, DEMOCRATIC
INTELLIGENCE EXCHANGE
While the days immediately after 9/11 may not have permitted reflection on the
proper safeguards for efficacious, humane arrests and interrogations, the problems
that have emerged from recent intelligence cooperation call for their institution.
184

This section proposes two mutually enforcing methods to reduce human rights
abuses and poor intelligence, support policies in favor of democracy and rights, and
reassert democratic control over intelligence activities. This section offers hope that
the combination of internal controls and democratic rules can allay some of the
excesses and irrationalities that arise from the participation of democratic states in
intelligence networks that often include repressive states.
179. Peter Baker, Old Enemies Enlist in U.S. Terror War: Former Soviet Republics Become Allies,
WASH. POST, Jan. 1, 2004, at A1.
180. Id.
181. GOODMAN & JINKS, supra note 60, at 8. Another danger, which Jinks and Goodman consider
an identifying characteristic of acculturation, is decoupling which involves adoption of structural
commitments that do not correspond to local practices such that true human rights compliance does not
occur. Id. at 3334. For example, Uzbekistan, a U.S. partner in the war on terror, now understands that
other states attach importance to human rights and accountability and has enacted legislation and brought
prosecutions against police for torture; yet, no conviction based on coerced confessions has been
overturned. Jones et al., supra note 68, at 78.
182. Another example of negative norm transmission is the pressure exerted through a strong lobby
of the foreign intelligence functionaries, who in order to maintain their influence successfully pressured to
maintain a post-Soviet model of the secret services in Poland, despite significant concern that such
services were persistent rights violators and extremely politicized. Andrezej Zybertowicz, An Unresolved
Game: The Role of the Intelligence Services in the Nascent Polish Democracy, in WHOS WATCHING THE
SPIES? ESTABLISHING INTELLIGENCE SERVICE ACCOUNTABILITY 145, 147 (Hans Born et al. eds., 2005).
183. This is confirmed by the network literature, which suggests that smaller or more-cohesive
networks exhibit a higher degree of internalization of and compliance with network standards. See, e.g.,
Slaughter, supra note 22, at 359.
184. STEPHEN GREY, GHOST PLANE: THE TRUE STORY OF THE CIA TORTURE PROGRAM 15152
(2006).
First, regulation internal to intelligence networks must be strengthened, as
Section A argues. Agencies purporting to represent the interests of democratic
governments should ascribe to more stringent and ethical professional norms for the
sake of accuracy, human dignity, and overall efficacy. They must use the
reputational sanctioning of network relations to acculturate repressive partners to
more ethical rules of the game. While there is substantial room for improvement of
intelligences democratic accountability, a number of networked intelligence services,
especially in the global counterterrorism effort, are not held to account in any way.
All intelligence agencies are, however, subject to reputational sanctioning from their
fellow professionals. Professional standards within the network, therefore, offer the
most promising basis for accountability.
Second, efforts internal to intelligence networks must be supplemented by
greater public and legislative involvement. Section B argues that the protection of
democratic governance and mitigation of intelligence excesses requires oversight,
accountability, and some measure of transparency. As they currently exist,
traditional means of accountability systematically fail to provide any of these.
Legislative oversight and public involvement are imperative to set proper limitations
and permissions on our intelligence services. Greater transparency in turn will
facilitate democracy and rule of law. Ultimately, as Section C contends, problems
related to the accountability and accuracy of intelligence may be best solved by
prioritizing law enforcement.
A. Acculturation of Intelligence Agencies to Ethical Professional Norms
If the experience of the last nine years is any indication, existing professional
standards in network arrangements are inadequate to ensure that intelligence
services act both in the most effective way and in the best interests of democracy.
Connections to untrustworthy, unreliable intelligence partners create a substantial
risk of bad foreign policy decisions, false positives, and subversion of democracy.
This section proposes ways to mitigate some of these problems. Section 1 argues in
favor of clear ethical professional standards to advance reliable intelligence sharing
and human rights compliance. Section 2 contends that the most effective way to
institute compliance with such standards is through acculturation within the
networkthat is, communitarian pressure on illiberal agencies to adopt more
professional, rights-respective behavior. Official training and aid, the quid pro quo
delivered by Western intelligence, must also support these efforts and the wider
foreign policy goals of democratic states. Ultimately, however, Western intelligence
may need to establish limits and cut or minimize ties to certain partners.
1. New Rules for Intelligence Cooperation
It is vital that information exchanged be precise and accurate and that the
means used to collect it be ethical and minimally intrusive on individual rights. This
section proposes clear, ethical professional standards to further information sharing
against transnational threatsin both the long and short term.
The first section contends that professional norms must include prohibitions on
torture and indefinite or arbitrary detention. The second section proposes
professional rules that make caveating information mandatory, rather than
discretionary. The third section argues in favor of use restrictions to ensure that
information sent to other states does not foster human rights abuses.
a. Ethical Standards in Compliance with International Human Rights
Law
Abuses generate inaccurate information, alienate populations, and impede
successful, rights-respecting cooperation. They do not constitute an effective
strategy to counter transnational threats. In order to deter rights violations and
mitigate their effects, professional intelligence standards should include principles
relating to the handling of intelligence and the arrest, detention, and interrogation of
suspects. There must be clear norms barring the use of coerced information. This
will avert Western complicity in human rights violations and incentivize placing
greater pressure on rights-violating partners to comply with proper intelligence
gathering techniques.
As a general rule of thumb, intelligence ethics should include principles of
justification, proportionality, necessity, and accuracy. All intelligence partners
should understand that individuals will be targeted only when it is justified,
authorised and the information gathered will be properly recorded and only retained
or made accessible where legitimate need can be established.
185
As the 9/11
Commission recommended, the United States should work with friends to develop
mutually agreed-on principles for the detention and humane treatment of captured
international terrorists . . . .
186
Such norms must bring the principle of harm
minimization to the forefront with the result that individual intelligence officials
confronting ethical dilemmas recognize them, seek guidance in codes of practice, and
ultimately minimize the potential harm to democracy, individuals, and consumers of
intelligence.
187
Like military professionals, who are governed largely by their
professional ethical rules, intelligence officials should take care not to act in such a
way that disregards the notion that individual human life and dignity are valuable for
their own sake and that people should be treated as an end in themselves and not
merely a means.
188

Professional intelligence services should abide by international prohibitions on
torture, summary execution, and cruel, inhuman, and degrading treatment. As the
Harvard University Project on Justice in Times of Transition has proposed in the
Intelligence Bill of Rights, agencies must conduct their work in a manner consistent
with respect for human rights as such standards are defined by international law,
including prohibitions on [t]orture and other [c]ruel, [i]nhuman or [d]egrading
[t]reatment or [p]unishment.
189
Agencies should further refrain from measures
185. GILL, supra note 2, at 153.
186. 9/11 COMMISSION REPORT: FINAL REPORT OF THE NATIONAL COMMISSION ON TERRORIST
ATTACKS UPON THE UNITED STATES 379 (2004) [hereinafter 9/11 COMMISSION REPORT].
187. GILL & PHYTHIAN, supra note 79, at 155 ([N]ot just individual security officials are moral
agents; so are the agencies and governments of which they are part, so statutes, guidelines and codes of
practice must all be drawn up within the context of ethical agreements . . . . Since intelligence cannot be
disinvented, and current practice is dominated by realist ethics, perhaps the most we can strive for is harm
minimization . . . .).
188. Pfaff, supra note 174, at 2.
189. Project on Justice in Times of Transition, Harvard University, Safeguarding Human Rights in
Relation to Intelligence Activities: An Intelligence Bill of Rights, para. 6 (Nov. 17, 2004) (on file with
designed to interfere with the normal political or judicial processes or with the lawful
internal workings of parties and organizations engaged in lawful activity, such as
spying on migr communities or peaceful protesters.
190

A commitment to refrain from exploiting information obtained through
coercive tactics should similarly travel in the toolbox of Western intelligence norms.
The determination of whether information has requisite reliability and credibility
should include an evaluation of whether the information was obtained through
torture or cruel, inhuman, or degrading treatment. If coercion has been used, the
information should be designated of dubious reliability. As recommended by the
Ottawa Principles on Anti-Terrorism and Human Rights, it should not be used as a
basis for:
(a) the deprivation of liberty;
(b) the transfer, through any means, of an individual from the custody of one
state to another;
(c) the designation of an individual as a person of interest, a security threat or a
terrorist or by any other description purporting to link that individual to terrorist
activities; or
(d) the deprivation of any other internationally protected human rights.
191

While this limit would not eliminate rights violations, it would curtail receiving
states ability to exploit those violations. This is in contrast to the current system
under which Western agencies encourage coercive tactics by gladly accepting
information gained through them. Over time, limits on use of coerced intelligence
might reduce torture by intelligence services. Just as the exclusionary rule serves as a
deterrent for the use of coercive tactics in the domestic context,
192
these restrictions
would bolster incentives for services to professionalize and train partners to engage
in more humane practices.
b. Caveats as a Requirement
Because intelligence is used to inform decision making on vital foreign and
domestic issues and may have serious consequences for individuals, it is imperative
that agencies institute clear requirements for quality control of shared information.
First of all, before information is shared, it ideally should be checked against other
available sources.
193
When information is shared with a partner, there is the implicit
expectation that the data you provide is accurate and that there are steps to ensure
author) [hereinafter Intelligence Bill of Rights] (proposing basic guidelines for intelligence based on a
study of intelligence systems and their regulation).
190. Id. paras. 67.
191. Ottawa Principles on Anti-Terrorism and Human Rights, Principle 8.1.2 (Oct. 16, 2006),
available at http://aix1.uottawa.ca/~cforcese/hrat/principles.pdf [hereinafter Ottawa Principles].
192. This is not a perfect analogy because intelligence agencies use information less often than law
enforcement agents. In the context of terrorism intelligence sharing, this rule would be stronger if more
criminal prosecutions for violent criminal acts were broughtas they indeed should be. See infra part
III.C.
193. New Challenges, supra note 51, para. 6 (It is vital that intelligence should be reliable, which
means sound intelligence that can be cross-checked against different sources. Complementary sources are
an operational requirement.).
information quality.
194
Second, proper caveating on shared intelligence should be
required, rather than discretionary. Lacking definitive guidance, agencies engage in
indiscriminate reporting of unverified information, without regard to the
information quality, reliability or usefulness, or without considering the receiving
agencys ability to analyze the informationgenerating a highly ineffective
information sharing environment.
195

The findings of the inquiry into Arars case amply demonstrate the importance
of proper caveating to ensure accuracy and minimize the negative repercussions of
false positives. According to the inquiry, as part of the RCMPs terrorism
intelligence responsibilities, it provided a great deal of misleading, inaccurate, and
piecemeal information to the United States. For example, the RCMP reported that
Arar and his wife were Islamic Extremist individuals suspected of being linked to
the Al Qaeda terrorist movement, had refused to be interviewed, and had
connections to known terrorists.
196
Officials systematically failed to attach written
caveats that indicated the reliability of the source and relevance to the
investigation.
197
This failure to caveat was not an isolated occurrence, but rather a
direct consequence of the informal and often implicit nature of understandings and
requirements between networked actors. By contract, requiring caveats could
protect individuals implicated, prevent unwise foreign policy decisions, and ensure
the integrity of intelligence sharing and collection processes.
198

In all cases, agencies should attach quality and reliability caveats. Each piece of
information shared should indicate whether it has been cross-checked and confirmed.
Reliability scales, contextual data, and the greatest information possible about the
source and method of collection should be indicated along with doubts or conflicting
data.
199
Codes for different sources could make information possible to track without
exposing sources identifying features, and meters of credibility could reduce circular
reporting and intelligence agencies intoxicating each other.
200
Since intelligence
agencies are charged with analyzing the reliability of information as accurately as
possible, the requirement that intelligence be caveated should not be overly
burdensome.
201

194. U.S. DEPT OF JUSTICE, PRIVACY AND INFORMATION QUALITY POLICY DEVELOPMENT FOR
THE JUSTICE DECISION MAKER 3 (Sept. 2005) (Promoting information quality by internal safeguards and
procedures helps to ensure the accuracy of the information you handle.). See also MARKLE FOUND.,
MOBILIZING INFORMATION TO PREVENT TERRORISM: ACCELERATING DEVELOPMENT OF A TRUSTED
INFORMATION SHARING ENVIRONMENT 23 (2006), available at www.markle.org/downloadable_assets/
2006_nstf_report3.pdf, [hereinafter MARKLE FOUNDATION] (If users do not believe that the information
is reliable and comprehensive . . . they will implement their own collection systems and may keep the
information to themselves.).
195. MARKLE FOUNDATION, supra note 194, at 19.
196. Melchers, supra note 99, at 4041.
197. Id.
198. Id.
199. MARKLE FOUNDATION, supra note 194, at 30 (arguing there also must be mechanisms to make
any limitations on the reliability or accuracy of data known to those using the information).
200. Mller-Wille, supra note 91, at 118119 (The exposure of intelligence products to a formalized
quality control would raise the demand for trackability of sources (codes can be used to prevent
exposure) and for clarity concerning what facts, assumptions and interpretations different conclusions are
based on.).
In the interest of accuracy, receiving agencies should only retain properly
caveated information that is accurate and complete.
202
Given the number of false
positives (and negatives) caused by mistaken identity and misspellings, a receiving
agency similarly should take care when combining information from various sources
that it all relates to the same individual.
203
At an institutional level, creating a
position for quality control might be helpful. Indeed, after the debacle over
allegations of weapons of mass destruction in Iraq, British MI6 created the position
of senior quality control officer to review and determine the credibility and veracity
of gathered intelligence.
204

In a similar vein, agencies should adopt rules to correct inaccurate
information.
205
This would involve a duty to investigate allegedly erroneous
information and to implement a process to correct inaccurate or materially unreliable
information.
206
Because of the transnational nature of intelligence cooperation,
correction of inaccurate information at the domestic level will not suffice. Informing
network partners upon the discovery of an error is crucial.
207
In the Arar case, for
example, this would have meant Canadian intelligence had a duty to notify both U.S.
and Syrian intelligence of the inaccuracy of information they had previously
forwarded.
c. Restrictions on the Use of Shared Information
Intelligence agencies should also attach use restrictions to prevent the use of
information to facilitate torture, arbitrary detention, or other abuses, as they already
do in other contexts. Some Western agencies screen intelligence for relevance and
personal information that, as a matter of domestic law, can only be shared in certain
circumstances and attach caveats limiting the informations uses.
208
Several of the
202. U.S. DEPT OF JUSTICE, PRIVACY, CIVIL RIGHTS, AND CIVIL LIBERTIES: POLICY TEMPLATES
FOR JUSTICE INFORMATION SYSTEMS C.3.10(c) (Sept. 2006), available at http://it.ojp.gov/
documents/Privacy_Civil_Rights_and_Civil_Liberties_Policy_Templates.pdf. See also Ottawa Principles,
supra note 191, at 8.2.2 (All state agencies involved in the collection and storage of personal information
must ensure that: (a) data are protected against unauthorized access, use or disclosure; (b) data are only
used in connection with the purpose for which they were collected; and (c) data are only held for as long as
necessary and are destroyed thereafter.).
203. U.S. DEPT OF JUSTICE, supra note 202, at C.3.10(d).
204. MARK M. LOWENTHAL, INTELLIGENCE: FROM SECRETS TO POLICY 31516 (4th ed. 2009).
205. Close allies may already tend to correct one another and withdraw inaccurate information. See
Senate Report on Iraqi National Congress, supra note 46, at 59, 167, 172 (providing examples of a foreign
agency withdrawing intelligence previously shared and indicating as an intelligence error the failure to
withdraw from circulation information shown to be unreliable).
206. MARKLE FOUNDATION, supra note 194, at 30; U.S. DEPT OF JUSTICE, supra note 202, at
C.3.10(e) (Investigate in a timely manner any alleged errors and correct or delete information found to
be erroneous . . . .); U.S. DEPT OF JUSTICE, supra note 194, at 7 (proposing as a central principle that
agencies [i]mplement safeguards to ensure information is accurate, complete, and current, and provide
methods to correct information discovered to be deficient or erroneous).
207. U.S. DEPT OF JUSTICE, supra note 202, at C.4.20 (When a participating agency gathers or
receives information that suggests that information originating from another agency may be erroneous,
may include incorrectly merged information, or lacks relevant context, the alleged error will be
communicated . . . .); Ottawa Principles, supra note 191, at 8.3.3 (States sharing information have an
obligation to correct information once they learn of its unreliability. States agencies and/or private
companies involved must be subject to shared, joint and several liability where errors or abuses occur.).
208. Arar Commission ReportAnalysis and Recommendations, supra note 46, at 18, 2223.
United States Western partners attach use restrictions to law enforcement
information, prohibiting the use of shared information to impose the death penalty.
209

Although there is a danger that use restrictions could chill intelligence sharing,
210

misuse of intelligence and abuses by Western intelligence services is likely to have a
similar impact on the willingness of agencies to share intelligence with those
services.
211
Cruel treatment at the hands of U.S. and other intelligence partners has
already prompted some Western intelligence agencies to consider restricting uses of
shared information as well. One of the Arar Commission recommendations was that
Canada evaluate sharing on a case-by-case basis and carefully limit potential uses of
shared information.
212
Canadian agencies are now required to consider limiting
cooperation with agencies in cases where individuals might be exposed to human
rights abuses or torture.
213
In other instances, partners have withheld information
unless they receive assurances it will not be used for proceedings in military
commissions or with relaxed evidentiary standards.
214
Still others have considered
limiting intelligence support to the United States out of fear that this relationship will
make them targets for attack.
215

To supplement use restrictions for individual pieces of intelligence, the
amendment of MOUs or formal agreements for the sharing of information may be
necessary. Such action would not be unprecedented. For example, when Israel used
U.S. satellite images to strike Iraqs Osirak reactor, against American intentions and
209. E.U. NETWORK OF INDEPENDENT EXPERTS IN FUNDAMENTAL RIGHTS, supra note 154, at 19
20 (noting that the first example was the refusal of France to cooperate with the United States in the case
of Zacarias Moussaoui until it received commitments that information provided by France would not be
used to impose the death sentence); Rudner, Hunters and Gatherers, supra note 10, at 214 ([A]llies like
Britain, France, Germany, and Spain have refused to extradite suspected al-Qaeda terrorists to the United
States, where they might face capital punishment.).
210. Melchers, supra note 99, at 42 ([S]ubsequent recommendations pertaining to the sharing of
investigative information and intelligence with foreign governments may provoke some degree of chill in
Canada-U.S. relations, though not unjustifiably so.); Rudner, Canadas Communications Security
Establishment, supra note 106, at 124 ([F]or Canada (or another partner country) to impose national legal
or human rights standards unilaterally onto Sigint interceptions might well jeopardize future UKUSA
collaboration against transnational crime and other sensitive targets.).
211. MARKLE FOUNDATION, supra note 194, at 24.
212. Arar Commission ReportAnalysis and Recommendations, supra note 46, at 348.
213. Melchers, supra note 99, at 43. The Canadian Security Intelligence Service has restricted its
cooperation with at least one foreign counterpart because of human rights concerns. Rudner, Hunters and
Gatherers, supra note 10, at 214.
214. AMERICAN COLLEGE OF TRIAL LAWYERS, REPORT ON MILITARY COMMISSIONS FOR THE
TRIAL OF TERRORISTS 2223 (March 2003) (noting that France and Germany have announced that they
will not disclose information to U.S. law enforcement authorities is that information could be used to
convict an alleged terrorist who might then be subject to the death penalty). See Walsh, supra note 37, at
62930 ([C]oncerns about civil and political rights might preclude one state from sharing intelligence with
another. Such states might be reluctant to share intelligence in their possession with receivers who have
weaker data protection laws or norms.); James Igoe Walsh, Intelligence-Sharing and United States
Counter-Terrorism Policy, in EMERGING TRANSNATIONAL (IN)SECURITY GOVERNANCE: A STATIST-
TRANSNATIONALIST APPROACH 44, 47 (Ersel Aydinli ed., 2010) (Governments have legitimate reasons
to surround their intelligence gathering and analysis with considerable security.).
215. Simon Tisdall, Fighting Terror the Malaysian Way, Not the US Way, GUARDIAN (U.K.), June 8,
2005, at 16 (reporting that Malaysia was focused on not becoming a target and was only cooperating
quietly); see also ALFRED B. PRADOS, CONG. RESEARCH SERV., IB 93085, JORDAN: U.S. RELATIONS
AND BILATERAL ISSUES 8 (2006), available at http://www.fas.org/sgp/crs/mideast/IB93085.pdf (Jordans
cooperative relationship with the United States has made it vulnerable to terrorist attacks, particularly
from organizations operating from Iraq.).
interests, U.S. intelligence amended its agreement with Israeli intelligence so as to
limit Israels use of U.S. intelligence to defensive purposes only.
216
At the domestic
level, several proposals have suggested use caveats or authorized use principles as a
way of facilitating information sharing.
217

Respect for these information sharing standards, which promote accuracy and
ethical intelligence behavior, can be developed in two reinforcing ways. First,
pressure from well-respected network partners can help acculturate illiberal agencies
to professional and rights-respecting norms, as Section 2 will discuss. Second, as
Section B will argue, standard setting and oversight through democratic bodies can
reinforce network acculturation to ethical intelligence standards and encourage
intelligence behavior in the interest of liberal democracies.
2. Professionalization of Illiberal Agencies
To the extent that Western intelligence must cooperate with services in illiberal
regimes, it is in the interest of all network partners that illiberal agencies be
professionalized and acculturated to professional norms through reputational
sanctioning within the networks. Unless the network acculturates less reputable
agencies to accept accurate, reliable information gathering skills and safeguards for
the humane treatment of suspects, it will be unable to provide consistently useful
informationespecially in the long-term. Successful transnational cooperation may
necessitate maintaining contact with certain repressive regimes, but consistent
pressure should be exerted on partners to improve their practices and become more
professional.
Professional standards, rather than legislative or other domestic oversight, are
the prime mechanism to sanction and professionalize network partners.
218
Therefore,
the use of peer accountability and reputational sanctioning to enforce ethical
professional norms, including prohibitions on mistreatment, may present the most
effective mechanism to professionalize repressive intelligence agencies.
219
Changing
the culture of intelligence agencies is without a doubt a difficult task, and [l]egal and
ethical standards have to be taken seriously if they are to become part of the
organizational culture, rather than just window dressing.
220
Effective acculturation
to ethical professional standards will require concerted reputational sanctioning and
targeted use of intelligence aid and training. If leading intelligence powers set
216. Lefebvre, supra note 4, at 536.
217. U.S. DEPT OF JUSTICE, supra note 194, at 7 (setting forth a use limitation principle, which would
require purpose specification and subsequent use only in conformance with such purposes); MARKLE
FOUNDATION, supra note 194, at 35 (proposing an authorized use criterion which must be articulated and
recorded prior to sharing).
218. Ian Cameron, Beyond the Nation State: The Influence of the European Court of Human Rights
on Intelligence Accountability, in WHOS WATCHING THE SPIES? ESTABLISHING INTELLIGENCE SERVICE
ACCOUNTABILITY 34, 41 (Hans Born et al. eds., 2005) (It is probably correct to say that a system of
internal controls, in particular the maintenance of the democratic sensibilities of the staff themselves, is
the most important safeguard against abuse of power.).
219. Id.
220. GILL & PHYTHIAN, supra note 79, at 15758.
standards through intelligence collaboration, restraint, and objective assessment,
other professional norms can spread to less professionalized agencies.
221

This section recommends using professional sanctioning and transmitting norms
through intelligence networks to encourage more effective long-term
counterterrorism information sharing and overall strategy. It relies primarily on the
theory of acculturation and argues that where hegemons or groups of influential
intelligence agencies encourage compliance with certain norms, other intelligence
agencies may adopt and comply with them. The idea of communitarian pressure
operating within intelligence networks to induce human rights compliance may seem
unlikely, but changes may occur, especially where well-respected intelligence
agencies target their support and sanction violators collectively and consistently.
222

In the past, Western intelligence powers exported their professional standards
and methods through networked contacts and transparent foreign aid. As Michael
Herman notes, because of the involvement of the U.S. and U.K. in the development
of various intelligence agencies worldwide, [i]ntelligence to some extent has its own
international patterns. Different national structures have points in common, with
influence and imitation operating on transnational intelligence networks.
223

Concepts beyond immediate operational capabilities have long been prominent in
Western attempts to acculturate and train foreign partners. Assistance from Britain
has focused on training to improve the objectivity of threat analysis, especially on
the civilian side of government and strengthening the capacity of intelligence
services to assess genuine outside threats with the understanding that depoliticized,
objective intelligence figures among the attributes of good governance and
responsible international citizenship.
224
Similarly, after the Cold War, the U.K. and
U.S. both focused on building the democratic accountability of intelligence services
in Central Europe.
225

Such use of intelligence networks suggests the capacity of intelligence contacts
to provide a lever for change.
226
Inclusion in a network can lead to pressure to
imitate and identify with the group, which is likely to generate compliance with the
groups norms.
227
Within intelligence networks, interactions among partners and
concern over the development of reputation can drive compliance with professional
norms.
228
In order to accomplish this, the network must provide clearly identifiable
221. See Herman, supra note 5, at 238.
222. HERMAN, supra note 131, at 219 ([T]he OECD nations plus some others signed a bribery
convention in which the United States has got all the right countries to play by roughly the same rules.
This is still far removed from intelligence; but it is still a reminder that unexpected things can happen when
states are persuaded of common interests.); Herman, supra note 5, at 238 (noting that during the Cold
War, the U.S. and USSR agreed not to extend some of their encryption for the mutual benefit of
facilitating verification).
224. Herman, supra note 5, at 230 (quoting the British Secretary of State for International
Development). Herman suggests that [p]erhaps Canada, Australia and New Zealand could make
particular contributions where US and UK advice is suspect, as perhaps could Germany whose post-1945
intelligence has fewer associations than most with covert activities. Id. at 237.
225. Herman, supra note 5, at 237.
226. Jones et al., supra note 68, at 18.
227. See id. at 34 (suggesting that the difficulty of reforming El Salvadors security forces after the
human rights tragedies of the civil war is an example of this pressure to imitate group norms).
228. Raustalia, supra note 8, at 51 (concluding that when networks promote regulatory change,
change occurs more through persuasion than command).
membership benefits.
229
For less reputable agencies, the anticipated long-term
benefits of a good professional reputation may then outweigh the present value of
violating network standards. Rewarding good behavior similarly may help induce
compliance by less reputable intelligence agencies. Rewards must, however, be
carefully constructed and include incentives for progress toward greater respect for
democracy and human rights.
Agencies should also use the capacity of intelligence networks to more
consistently penalize violations of rules. Intelligence professionals are those most
likely to become aware of impropriety by partner agents and are therefore the best
actors to conduct oversight and demand compliance with professional guidelines.
230

Clear rules, sanctions, and rewards facilitate efficacious intelligence sharing and
create predictable and consistent standards of professional behavior.
231
For, [i]f
there is no expectation that misuse of the system will result in a penalty, there is little
disincentive for misuse.
232

Maintaining full contact with a partner who continually violates professional
norms may impede sharing and trust-building. Communicating an agencys violation
of network standards, by contrast, can shame the agency and incentivize replication
of more professional, well-respected agencies in the network.
233
When a partner
agency fails to make progress toward professional and accountable behavior, several
alternative responses are possible: [a]t the lowest level, the system terminates
access by an offending user. If the problem is more pervasive in the agency, it can
terminate access by the agency. Finally, there is the option of maintaining access but
working with the agency to improve its practices and compliance.
234
Partner
agencies might similarly only share information with violating partners on a case-by-
case basis or in limited areas less likely to result in repression such as countering
biological weapons.
The goal of intelligence network sanctions and rewards should be to cultivate
capable, accountable, and professionalized network partners who can be trusted to
gather, analyze, and use information in humane and professional ways. As is the
aspiration of Western intelligence generally, intelligence networks should employ
reputational sanctions and aid in order to keep[] the players honest, not permit[]
disreputable arguments to thrive, [and] point[] out where positions are internally
contradictory or rest on tortured readings of the evidence.
235
The combination of
network sanctioning and targeted support can lead to greater professionalization of
our network partners. Regarding all contact with less reputable agencies, Western
229. LTG. PETER A. KIND & J. KATHARINE BURTON, INSTITUTE FOR DEFENSE ANALYSES,
INFORMATION SHARING AND COLLABORATION BUSINESS PLAN 52 (2005).
230. GILL & PHYTHIAN, supra note 79, at 158 ([T]here is increasing recognition of the need for some
international oversight mechanism to reinforce changes made to operational guidelines and training).
231. U.S. GOVT ACCOUNTABILITY OFFICE, GAO-02-04, INFORMATION SHARING: PRACTICES
THAT CAN BENEFIT CRITICAL INFRASTRUCTURE PROTECTION 8 (2001) (concluding that there must be
procedures for handling rules violations because a violation of partners trust will undermine purpose and
diminish willingness to share information).
232. MARKLE FOUNDATION, supra note 194, at 29 ([A] clear, calibrated, and predictable system of
accountability for misuse of the system should be an important part of the information sharing
environment . . . . Penalties should be known, and they must be applied consistently.).
233. GOODMAN & JINKS, supra note 60, at 7.
234. U.S. DEPT OF JUSTICE, supra note 202, at C.5.20.
235. HERMAN, supra note 9, at 152 (quoting R. Jarvis).
intelligence should seek to minimize potential harm and foster greater accountability
and respect for professional standards. In contrast to current practice, Western
intelligence should recognize when the costs outweigh the benefits and when
withdrawal of support is in order. Ultimately, raising the professional level of
partners through training, aid, and support is imperative. The next section will
examine the most effective way to attain this goal.
3. Limits on Quid Pro Quo
This section argues that training and aid should be tailored to further
professionalization, and that the quid pro quo for counterterrorism support should
minimize potential contribution to rights violations. It relies on studies that show
that professionalization through training, aid, and network contact of security
services can enhance accountability and efficiency of security services.
To foster greater respect for accuracy, depoliticization, and proposed
prohibitions on cruel treatment and arbitrary detention, it is imperative that Western
nations seriously consider the most appropriate and effective form of assistance to
less professional agencies. Future aid should specifically take into account the short-
and long-term costs of counterterrorism efforts. Although long-term costs are
frequently ignored, counterterrorism efforts have substantial short-term costs as well.
For example, not only did U.S. support to Pakistani intelligence and security forces
feed anti-American sentiment and religious extremism there (a long-term cost), it
also caused the Pakistani economy to lose in excess of $10 billion in two short
years.
236

Assistance to intelligence services should encourage progress toward
competence and accountability. Studies show that properly designed training and aid
can enhance the accountability and efficiency of more illiberal security services.
237
In
El Salvador, for example, qualitative and quantitative evidence reveal the success of
UN and U.S. efforts to improve the security services.
238
In less than a decade,
security services that had been responsible for thousands of summary executions and
systematic torture were called to account and made subject to external review and
human rights vetting.
239
The same study also found that aid and training of security
intelligence forces generally does not work in highly authoritarian countries, but can
have significant effects in emerging democracies.
240
Such findings should be a key
consideration when and where the United States decides to provide aid. All aid
should be evaluated on the basis of its effects on human rights as well as its
contribution to security.
To avoid contributing tools to its repressive partners that facilitate human rights
abuses or subvert democracy, Western intelligence should focus on areas with less
236. K. ALAN KRONSTADT, CONG. RESEARCH SERV. IB 94041, PAKISTAN-U.S. ANTI-TERRORISM
COOPERATION 4, 6, 17 (2003), available at http://fpc.state.gov/documents/organization/27323.pdf.
237. See, e.g., Jones et al., supra note 68, at xiixvi, 6 (conducting a multi-country study of U.S. aid to
internal security forces in Pakistan, Uzbekistan, El Salvador, and Afghanistan and concluding that there is
evidence that providing assistance to internal security forces can lay the groundwork for future support for
reform to human rights practices).
238. Id. at 3841.
239. Id.
240. Id. at xixiii.
risk of rights violations. Some assistancesuch as bomb-squad training and counter-
proliferation supportmay increase security with little impact on rights, and thus
provide a mechanism for continued intelligence contact and the promotion of mutual
interests.
241
Other apparently innocuous techniques can create the potential for
further repression.
242
Where possible, money should be targeted to certain specific
benchmarks, because due to its fungibility it carries risks of undermining foreign
policy and reform.
243
In all cases, assistance should be assiduously conditioned, and
further aid should depend on progress made toward improving practices.
Training must consistently impart the lesson that arbitrary arrests, detention,
and torture are unacceptable. In the United States, in the late 1970s, executive policy
mandated just such training in a move that was welcomed by many within the
intelligence community.
244
Jack Devine, the former acting head of the CIAs
worldwide operations observed that the refusal to descend to the level of repressive
and Communist regimes gave the United States an ideological advantage.
245

Discarded in the heady post-9/11 days, these standards gave credence to the principle
that Western intelligence agencies acting as an arm of democratic governments
would uphold, not undermine, the values of liberal democracies. Since 9/11, U.S. and
other Western intelligence agencies have wandered far from the promotion of these
values and have faced significant dilution of professional norms and damage to their
reputations.
Ultimately, Western intelligence agencies should know when to cut ties with
certain partners. While they should not expect perfect compliance with professional
standards and rights norms to result from these efforts, they must weigh the costs of
associating with a service that continues to commit abuses. Agencies from liberal
democracies should consider the danger of bolstering the capacity of their partners
that carry out the policies of brutal authoritarian governments.
246
Certainly, an
agency should end, reduce, or significantly restructure assistance, when it bears no
fruit.
247

B. Establishing Democratic Oversight and Accountability
This section argues in favor of more watchful and involved legislative and
transnational bodies, engaged public debate, and a greater degree of transparency.
So doing, it sets out the general principles that should underlie proposed domestic
and international mechanisms. Fundamentally, there should be more comprehensive
and transparent debate over the purpose, guidelines, and limits of transnational
intelligence networks. Because the tension between the demands of democracy and
241. Id. at xiv.
242. Id. at 87 ([E]ven forensic training (including training in explosives) raises concerns that it could
enable Uzbek personnel to more effectively fabricate evidence in criminal and counterterrorism
proceedings.).
243. See, e.g., Van Natta, supra note 173, at A22 (reporting that the U.S. State Department cut off $18
million in aid because the Uzbek government failed to improve its human rights record, but the U.S.
military then provided an additional $21 million for removal of biological weapons).
244. GREY, supra note 184, at 13.
245. Id.
246. Jones et al., supra note 68, at xviiixix.
247. Id. at 85.
the secrecy of intelligence networks is particularly acute, the argument focuses on
democratic states with professional intelligence services.
Section 1 tackles the problem of the secrecy that obscures the activities of
transnational networks from public and legislative awareness. It proposes that
intelligence network agreements and MOUs be declassified and that, with the
passage of time, originator controls not be permitted to trump Freedom of
Information Act requests and declassification of documents. Section 2 calls for
democratic bodies to turn their attention to transnational networks. For too long,
legislatures and the public have been unaware of and uninvolved in the transnational
activities of agencies that purport to serve democratic interests, but often subvert
them. Clear statutory permissions and limitations should be devised by democratic
representatives. While not without cost, the involvement of the legislatures and the
public would convey substantial advantages for intelligence agencies in terms of
greater rationality, consistency, and accuracy.
Section 3 argues that, as a general matter, law enforcement networks offer
greater potential to counter terrorism in an accurate, reliable, and rights-respecting
way and should, therefore, be the focus of international cooperation. Intelligence
and law enforcement officials have long worked together to counter and prosecute
terrorist groups. Law enforcement intelligence must be prioritized, and any genuine
problems involved in prosecutions and information exchange should be the focus of
international and transnational cooperation.
1. Improve Transparency through Declassification
Greater transparency is essential to proper accountability, oversight, and
vigilance on both the domestic and international levels. Democratic governance,
which depends on the consent of the governed, requires the public to have some
knowledge of intelligence activities. Tension between secrecy and accountability is
inevitable, but democratic states have erred too far in the direction of secrecy,
especially with regard to transnational intelligence sharing arrangements.
248
This
Article proposes that sharing agreements be made public and that there be a
presumption in favor of declassifying information shared through networks,
especially as time progresses.
a. Declassification and Legislative Approval of Intelligence Sharing
Agreements
Disclosure of the details of intelligence-sharing agreements (what quid pro quo
is promised) is key to public scrutiny and accountability. Other regulatory and
enforcement agencies regularly publish their MOUs. It is not clear why formal
intelligence agreements, some of which have been in existence for half a century,
remain classified as top secret. As Richelson and Ball argue, with regard to the
UKUSA relationships, the citizens of the five democracies [should] be officially and
fully apprised of the nature and operations of these agencies, and of the
consequences (both beneficial and disadvantageous) of the international cooperative
248. RICHELSON & BALL, supra note 12, at 310 ([T]he conflict of interest between the requirements
of secrecy and the basis of democratic government itself has come to be reconciled in each of the UKUSA
countries too far in favour of secrecy.).
arrangements among themto the extent permitted by the genuine requirements of
national security.
249
Under existing rules, however, declassification of the
agreements would likely require the consent of all agencies party to them.
Additionally, despite practical difficulties, future intelligence sharing agreements that
permit the exchange of personal information should be reviewed by legislatures to
ensure their compliance with international and national data protection and human
rights legislation.
250

Some fear that making information about networks public would endanger
security by permitting our enemies to know too much about the quantity and quality
of intelligence and the methods of sharing. This concern, however, is largely
overstated since other nations intelligence services already know a great deal more
about our intelligence agencies and their information sharing arrangements than the
public does.
251
While the level of transparency characteristic of other government
agencies cannot be expected, some measure of transparency is important to
democratic accountability and well-functioning, accurate intelligence. Equipped with
the knowledge of the intelligence partners and the mechanisms of sharing, citizens
would be able to engage with intelligence policy.
252

b. Presumption in Favor of Declassification of Shared Information
Declassification of information shared through networks is central to improving
accountability for actions taken through network arrangements. Declassification
creates the potential for delayed scrutiny. As we have seen in the United States,
historians, intelligence experts, and the National Security Archives have made
valuable use of information declassified and released years later.
253

It is likely thatas with other government agenciesintelligence shared
through networks is significantly over-classified.
254
To remedy this, more precise,
uniform classification standards for national security information should be
developed to ensure that information only remains classified so long as there are
specific and articulable facts suggesting that disclosure of such information would
cause identifiable harm to national security or critical infrastructure and that harm
249. Id. at 309. See also Ottawa Principles, supra note 191, 7.5 (Confidentiality rules that apply to
information-sharing agreements between states may not take precedence over the right of citizens to
access information from their governments.).
250. Ottawa Principles, supra note 191, 8.3.56.
251. See Aldrich, supra note 2, at 5152 (suggesting that the United States has long maintained a
clandestine kinship with intelligence services in Syria and Libya, despite public condemnation of these
services during the 1990s for their association with terrorists).
252. Steven C. Boraz, Establishing Democratic Control of Intelligence in Colombia, 19 INTL J. OF
INTELLIGENCE & COUNTERINTELLIGENCE 84, 102 (2006) (arguing that intelligence accountability calls
for an engaged citizenry, public debate, training civilians to understand intelligence, and opening
intelligence training schools for those who engage in oversight).
253. See 2005 NATIONAL SECURITY ARCHIVE ANN. REP., at 111 (2006) (detailing many uses of
declassified information obtained by the National Security Archives).
254. U.S. GOVT ACCOUNTABILITY OFFICE, GAO-06-706, MANAGING SENSITIVE INFORMATION:
DOD CAN MORE EFFECTIVELY REDUCE THE RISK OF CLASSIFICATION ERRORS 12 (2006) ([F]ormer
DOD Deputy Under Secretary of Defense for Counterintelligence and Security . . . estimated that 50
percent of information may be overclassified, to include overclassification between the classification
levels.).
outweighs the public interest in disclosure.
255
Failure to declassify information
whose sources and methods no longer need protecting is irrational, as it responds to
no national security or foreign policy objectives.
Subject to the requirement that classification supports a legitimate national
security interest, a presumption in favor of declassification should be instituted and
gradually grow heavier over time. The interest in secrecy recedes with the passing
years as targets change, sources die, and methods become outdated.
256
Sometimes,
this can occur very quickly; for example, the release of information relating to
Saddam Husseins regime should raise no serious concerns, as the regime has been
toppled, and no one doubts that foreign intelligence agents were actively spying
there.
Under such circumstances, the consequences of declassification generally will be
beneficial, not harmful, to proper intelligence work. Intelligence operatives report
that the current systems presumption of secrecy makes it tempting to ignore
longer-term costs.
257
Transparency, therefore, can be expected to generate more
effective and efficient intelligence over time.
258
The knowledge that their activities
will be made known in the future should generate intelligence policies and behavior
that resonate more clearly with democratic values. For example, French intelligence
agents might have reconsidered engaging in surveillance of peaceful Iraqi dissident
groups in France and providing reports to Iraqi intelligence had there been a system
that would eventually declassify their wrongdoing. Instead, this collusion, which the
French people did not support, was revealed by documents seized during the Iraq
invasion.
259

Indeed, wrongdoing often eventually emerges to the detriment of an
intelligence services reputation and respectthrough media reports, whistleblowers,
or eventual declassification.
260
The result is that policymakers and other intelligence
consumers question the judgments of intelligence agencies and doubt their
255. Christina E. Wells, National Security Information and the Freedom of Information Act, 56
ADMIN. L. REV. 1195, 121820 (2004); see also ARTICLE 9, JOHANNESBURG PRINCIPLES ON NATIONAL
SECURITY, FREEDOM OF EXPRESSION AND ACCESS TO INFORMATION 1113 (1996), available at
http://www.article19.org/pdfs/standards/joburgprinciples.pdf (setting forth general rule on access to
information and narrow use of security exemption so as to further public interest in disclosure of
government information).
256. Mller-Wille, supra note 91, at 12122 (The issue of releasing intelligence is merely a matter of
determining the date for declassification, not a matter of principle and national security. The argument of
the necessity to protect sources and methods disappears with time.); Thomas Blanton, The Worlds Right
to Know, FOREIGN POLY 50, 56 (Jul./Aug. 2002) (Former U.S. Secretary of State Lawrence Eagleburger
has said most of the secrets he saw in his government career could easily be released within 10 years of
their creation.).
257. TURNER, supra note 41, at 139 (citing veteran CIA operative Gregory Treverton); see also
Bruneau & Dombroski, supra note 31, at 163 (If the intelligence agencies know that in the future their
files will be open for public scrutiny, they are logically more likely to keep a rein on the behavior of their
members.); RICHELSON & BALL, supra note 12, at 310 (Secrecy has shielded these agencies from full
accountability and effective supervision and led to their being less effective and less efficient than they
otherwise might be.).
258. Blanton, supra note 256, at 52 (noting that openness fights terrorism by empowering
citizens . . . and holding officials accountable).
259. Rudner, Hunters and Gatherers, supra note 10, at 220.
260. RICHELSON & BALL, supra note 12, at 310 (The truth inevitably prevails, and operations which
have been undertaken on the premise that they could be plausibly denied will in the end only damage the
reputation of the UKUSA countries and the faith of their citizens in their governments.).
legitimacy.
261
What is reported and leaked regarding intelligence largely brings
transgressions and errors to the public attention with a negative effect on
motivation and morale among intelligence agents.
262
By contrast, greater
transparency through declassification of information should help bring intelligence
successes as well as failures to light and result in greater legitimacy for intelligence
agencies.
2. Establish Democratic Control and Oversight
Hand in hand with greater transparency must come more rigorous oversight.
While the precise features of oversight are specific to the historical and institutional
experience of each country, it is clear that current oversight mechanisms in every
nation must be rethought. The principle of national controlcompelled by
democratic accountabilityshould guide the development of new mechanisms. To
counter agencies temptation to use networks to subvert democratically determined
policy, standards of legality and propriety of intelligence cooperation must be
decided and enforced at the national level.
263
Each democratic state should
institutionalize a regular reassessment of the costs and benefits of intelligence
relationships with its partners.
264
Decisions should be made by elected officials,
instead of a closed group of professionals who are largely insulated from the
demands of their constituent publics or, worse, a group of foreign intelligence
officials who make changes in policy through their influence on domestic intelligence
services.
265
Such oversight mechanisms will contribute to, rather than detract from,
the effectiveness of intelligence services and ensure that engagement in transnational
networks advances the protection and promotion of democracy.
Clear statutory permissions and limitations, as Section a sets out, are imperative
to proper oversight and democratic involvement in the transnational activities and
connections of agencies. So too is budgetary control, as Section b argues.
a. Set Clear Statutory Authority and Limitations on Intelligence
Cooperation
The lack of democratic involvement in setting clear statutory permissions and
limitations must be remedied. The rule of law demands that no intelligence agency
be established in secrecy and that its legislative mandate, limitations, funding
261. Liaropoulos, supra note 2, at 1112.
262. LAQUEUR, supra note 90, at 231 ([B]ecause the work of intelligence has always proceeded at
least somewhat in the shadows, open recognition is seldom heard . . . there has been little encouragement
and many public attacks for U.S. intelligence.).
263. GILL & PHYTHIAN, supra note 79, at 152 ([A] legal framework for security intelligence is a
necessary, but not a sufficient, condition for democratic oversight.). See also Ottawa Principles, supra
note 191, at 9.1 (setting forth a monitoring regime for security intelligence activities that focuses on the
effectiveness, propriety, legitimacy, and accountability of intelligence activities).
264. RICHELSON & BALL, supra note 12, at 307 (As the Australian Royal Commission on
Intelligence and Security found in 1977, We . . . need constantly to re-assess the benefits to Australia from
intelligence relationships with other countries against the costs.).
265. JOHNSON, supra note 35, at 95.
sources, and oversight mechanisms be public.
266
As compared to professional
standards that currently regulate transnational networks, legislation and formal
democratically approved agreements have the advantage of clarity and precision.
They would focus expectations and set forth proscribed, permitted, and required
behavior.
267

Through legislation, domestic legislatures could play a significant role in better
evaluating the costs and benefits of intelligence agencies associations. They could
ensure that intelligence sharing networks respect basic human values and form part
of long-term, well-functioning strategy to deal with transnational threats. Such
legislative measures have been enacted in the past, but have suffered from
inconsistent application and fallen into desuetude. In the 1960s and 1970s, the CIAs
training of foreign partners in countersubversion, counterguerilla, and intelligence
gathering techniques often increased these foreign entities capacity for repression.
268

The U.S. Congress responded with a prohibition on the provision of U.S. funds and
support to any internal security service, police, or law enforcement.
269
This law likely
went too far. Some support, especially to security services in emerging democracies,
can improve the human rights practices of those services. Nevertheless, the brutality
of many current intelligence partners calls for legislative involvement to minimize
potential complicity in rights violations and to ensure effective foreign policy.
Legislative dialogue would take into account the panoply of foreign policy goals and
would consider international and national legal limits. It would therefore be a
promising starting point to reduce intelligence networks potential for complicity in
violation of rights and repression of democracy.
Legislatures in liberal states should require intelligence agencies to vet their
partners, weigh the potential for abuse against a clearly articulated benefit of
cooperation, and advance professionalism and humane treatment through their
support to foreign counterparts. The United States requires vetting to prohibit
funding or association with known human rights violators, but applies the policy
inconsistently; the policy also does not apply to the CIAs funding of foreign
intelligence partnersleaving a gaping hole in enforcement.
270
By establishing
266. Intelligence Bill of Rights, supra note 189, para. 2 (Intelligence services must be established
pursuant to duly enacted legislation by a democratically elected government. The establishment of the
agency, its mandate, its funding sources and the nature of its oversight must be made part of the public
record.).
267. See BIRKINSHAW, supra note 69, at 33 (observing that Canadas Commission of Inquiry on the
Royal Canadian Mounted Police concluded that [t]he agencys activities in relation to security and its
responsibilities should be defined in statute and not diffuse and ambiguous sources arising as they did
from a melange of Cabinet directives, ministerial correspondence and unstated RCMP assumptions).
Within Western democracies, these should include measures to ensure appropriate levels of privacy
protection not be undermined by transnational networks. See Ottawa Principles, supra note 191, at 8.2.5,
8.3; U.S. DEPT OF JUSTICE, supra note 194, at 34 (discussing importance of agents adherence to and
understanding of privacy restrictions).
268. Jones et al., supra note 68, at 11.
269. Id. at 11. Though still in effect, this law has been rendered toothless by widespread waivers and
exemptions. Id.
270. Id. at 15; see also U.S. GOVT ACCOUNTABILITY OFFICE, GAO-06-850, SECURITY ASSISTANCE:
LAPSES IN HUMAN RIGHTS SCREENING IN NORTH AFRICAN COUNTRIES INDICATE NEED FOR FURTHER
OVERSIGHT (2006) (concluding that human rights vetting was not done or was insufficient when
counterterrorism training and equipment were provided to security forces in Algeria, Morocco, and
Tunisia); U.S. GOVT ACCOUNTABILITY OFFICE, GAO-05-793, SOUTHEAST ASIA: BETTER HUMAN
RIGHTS REVIEWS AND STRATEGIC PLANNING NEEDED FOR U.S. ASSISTANCE TO FOREIGN SECURITY
proper vetting procedures, legislatures could set uniform guidelines and clarify the
criteria agencies should use to evaluate partners.
271
Specific to intelligence sharing,
legislatures in democratic states should seriously consider requiring, as Canada does,
a case-by-case evaluation and a finding that the disclosure to a service of a
government that violates rights is justified by Canadian security or law enforcement
interests . . . can be controlled by specific terms and conditions; and does not have a
negative human rights connotation.
272
With consistent national oversight and
statutory control, some of the excesses of involvement in transnational intelligence
networks might be reduced.
Domestic oversight mechanisms will be insufficient to counter all the abuses
and inefficiencies in intelligence networks. These oversight mechanisms will be
biased toward the protection of citizens, sometimes to the detriment of the interests
of non-citizens affected by the activities of intelligence agencies. They also will not
exist in the many states without robust democratic systems, with the result that their
agencies international cooperative arrangements will go unchecked.
These oversight deficits may call for an international or transnational
monitoring body. Slaughter suggests that once aware of networks of government
agencies, legislators will expand their oversight capacities and develop networks of
their own.
273
Within close-knit arrangements like the UKUSA network, some
cooperative oversight body might be plausible even if subject to more executive than
legislative control. It certainly would present an interesting and possibly more
effective way to regulate international networks.
274

Already, some small measure of contact between legislatures has resulted from
increasing awareness of intelligence agency network relations. For example, the
U.K. oversight body, the Intelligence and Security Committee, takes part in
international liaison and exchanges, both by visiting oversight agencies abroad and
receiving such visits (these have included many European and former Eastern bloc
countries, the United States, and the other Commonwealth states); it also conducts
many visits a year to agencies premises to examine their functioning.
275
Similarly, in
response to requests from Poland and Argentina during their transitions to
democracy, the U.S. Senate Select Committee on Intelligence provided assistance in
creating mechanisms for legislative oversight and direction of intelligence agencies,
including sending staff to those countries.
276
The Church Commission and U.S.
establishment of intelligence agency oversight also influenced legislatures in other
FORCES (2005) (concluding same with regard to Philippines, Indonesia, and Thailand).
271. Jones et al., supra note 68, at 173 (Congress can play a critical role by seeking to establish
uniform guidelines [for vetting practices] and providing further clarification regarding the criteria
executive branch agencies should use in identifying and vetting both units and individuals.).
272. COMMN OF INQUIRY INTO THE ACTIONS OF CANADIAN OFFICIALS IN RELATION TO MAHER
ARAR, REPORT OF THE EVENTS RELATING TO MAHER ARAR: FACTUAL BACKGROUND 24647 (2006)
(paraphrasing the RCMP Operational Manual).
273. Anne-Marie Slaughter, The Real New World Order, FOREIGN AFF., Sept.Oct. 1997, at 183, 197.
274. Aldrich, supra note 2, at 53. See generally Post Cold-War International Security Threats:
Terrorism, Drugs, and Organized Crime Symposium Transcript, 21 MICH. J. INTL L. 655 (1999) (arguing
that an international terrorism court run by the agencies of the G-8 would benefit intelligence sharing and
enable coordination of standards and procedures).
275. Leigh, supra note 98, at 90.
276. Searching for Answers: US Intelligence After September 11: A Conversation with Bob Graham,
HARV. INTL REV., Fall 2002, at 40, 43.
countries to move toward increased regulation and oversight of their intelligence
services.
277

Especially within the European Union, there are calls for harmonization and
more transnational or regional oversight of intelligence sharing and cooperative
networks. In light of the ECHELON investigation, the European Parliament
appealed to member states to draw up a common code of conduct establishing
proper levels of protection against intelligence operations that would bind each
national agency as to operations against anyone within the European Union.
278
It
further advocated including the United States in a common code so as to provide
maximum protection for EU citizens.
279
Within the European context, Bjrn Mller-
Wille suggests a model building on an independent oversight body composed of
professional experts.
280
As he envisions it:
[E]xperts, e.g., from national oversight bodies, could be appointed by
Member States and representatives from the [European Parliament] . . . EU
oversight bodies should report to their national equivalents (and through
them to national parliaments), the Council and, if not to the responsible
committee of the EP as a whole, at least to the five MEPs with security
clearance . . . . [T]hese bodies should be responsible for ensuring that
agencies act within their mandate and that they do not violate civil liberties
and constitutional rights.
281

Some form of transnational oversight, whether expert, executive or legislative,
may ultimately be necessary in order to solve some of the inconsistencies and
dangers particular to intelligence networks. Pre-existing networks, such as the
UKUSA agreements, or regional institutions, such as the European Union, might
provide a foundation for the development of such mechanisms.
b. Develop Budgetary Control
Key to democratic oversight and the prevention of intelligence agency
subversion of statutory or policy limits is budgetary control. Participation in
intelligence networks offers numerous benefits, but comes at a cost to democratic
rule, especially when agencies receive or impart funds without democratic
involvement.
Networks can prevent legislatures from exercising fiscal accountability, which is
a central and, in many countries, constitutionally-mandated tool to check the
executive.
282
Imagine a democratic country that attempted to limit the power of its
277. See Farson, supra note 100, at 226 (stating that within the UKUSA network, it was not until
attitudes shifted in the US, by then the alliances leading partner, that reforms occurred elsewhere).
278. EUROPEAN PARLIAMENT RESOLUTION ON ECHELON, supra note 93, para. 13. See also
Ottawa Principles, supra note 191, at 8.3.7 (The UN member states should develop and adopt an
international instrument affirming privacy and data protection as fundamental human rights and laying
down minimum standards for protection in accordance with these principles.).
279. EUROPEAN PARLIAMENT RESOLUTION ON ECHELON, supra note 93, para. 14.
281. Id.
282. Keohane, supra note 76, at 1132 (Fiscal accountability describes mechanisms through which
funding agencies can demand reports from, and ultimately sanction, agencies that are recipients of
funding.); 9/11 COMMISSION REPORT, supra 186, at 410 (When Congress passes an appropriations bill to
allocate money to intelligence agencies, most of their funding is hidden in the Defense Department in
national intelligence agency by restricting funding or issues on which it can gather
intelligence. Through networking with partners, however, an agency could escape
these funding limitsexpanding its available resources and areas of interest. Even
more problematically, the intelligence service could receive money or equipment in
return for network cooperation, which if kept from legislators could allow it to
become self-funded and more powerful than the people, through their legislators,
intended.
This may seem far-fetched, but it is not unprecedented. For instance, until the
1990s, ninety percent of the budgets of the Norwegian Secret Services came from the
United States and NATO without public or parliamentary knowledge.
283
Similarly,
the National Reconnaissance Office, a U.S. military intelligence agency, whose
existence was only publicly disclosed in 1994, accumulated a $4 billion slush fund of
appropriations without congressional knowledge.
284

Oversight mechanisms should therefore take into account funds received from
partners. With regard to large intelligence agencies which sometimes fund extremely
repressive partners, national overseers should be aware of how appropriations are
being dispensed. Under the current system, many countries budgets obscure the
most basic facts regarding the amount of funds available and how they are spent;
intelligence funds often simply disappear in the defense budget.
285
In the United
States, the 9/11 Commission advised Congress to pass a separate appropriations act
for intelligence, defending the broad allocation of how these tens of billions of
dollars have been assigned among the varieties of intelligence work.
286
In this way,
overseers could limit improper funding of illiberal regimes.
Budgetary control can be remarkably successful. As the former U.S. Director
of Central Intelligence Colby remarked, in order to convince the CIA to abandon a
plan, an [Intelligence] Committee chairman needs only to say to the DCI at the end
of the briefing: Write down in your notebook $100 million, becauseif you go
aheadthat is what is coming out of your CIA budget next year.
287

C. Rely on and Bolster Law Enforcement Intelligence
Due to the persistent human rights and democratic accountability problems
manifested by intelligence networks, the best way to counter transnational crime,
ensure accurate information, and reduce rights violations is to improve information
order to keep intelligence spending secret.). James Bamford has been quoted as saying that [b]ecause
the issue is hidden under heavy layers of secrecy, it is impossible for even Congress to get accurate figures
on just how much money [is being spent] and how many people are involved. Pratap Chatterjee, Military
Interrogation Training Gets Privatized, CORPWATCH.ORG, (Mar. 7, 2007), http://www.corpwatch.org/
article.php?id=11940.
283. Fredrik Sejersted, Intelligence and Accountability in a State Without Enemies: The Case of
Norway, in WHOS WATCHING THE SPIES? ESTABLISHING INTELLIGENCE SERVICE ACCOUNTABILITY
119, 121 (Hans Born et al. eds., 2005).
284. JOHNSON, supra note 6, at 103.
286. 9/11 COMMISSION REPORT, supra 186, at 416.
287. JOHNSON, supra note 35, at 135 (quoting DCI William Colby); accord Mller-Wille, supra note
91, at 113 (At the national level, budgetary control represents one of the most powerful means by which
parliaments can influence and sanction the executives policies.).
sharing networks for law enforcement and criminal convictions. Treating terrorist
violence as a criminal actto be handled through legal systems in accordance with
the rule of law, democracy, and human rightsrepresents the most consistent and
successful way to reduce violence and terrorist threats. Arrest and prosecution of
suspects, instead of detention and torture at the hands of less reputable partners,
would check abuses and engender public support. It would also involve the single
most valuable tool to discern inaccuracies, mistakes, and manipulation: an
adversarial process before an independent court.
1. Improve Law Enforcement Sharing Networks
Because of the benefits inherent in criminal prosecutions, the focus of the
international community and national legislatures should be on improving the law
enforcement networks, which link police, prosecutors, and judges.
With regard to transnational threats and actors, the primary approach has long
been law enforcement. Terrorist organizations generally rely on criminal acts for
their financing; criminal prosecution for lesser crimes, therefore, offers a mechanism
to disrupt terrorist activities and turn would-be terrorists into witnesses against
higher-ups.
288
Successful investigation and prosecutions of these crimes have long
demanded regular contact among law enforcement agents.
289
As a result, police and
law enforcement intelligence agencies typically have relied on close and often long-
standing connections to their counterpart agencies to counter transnational threats.
290

Only after the end of the Cold War did intelligence professionals, in an attempt to
remain relevant, begin to take over the field.
291

Prior to 9/11, law enforcement was effective against suspected terrorism.
Intelligence and law enforcement officials jointly operated against transnational
crime, including terrorism. U.S. intelligence and law enforcement (primarily the CIA
and FBI) routinely targeted leaders of terrorist groups together. In the context of
their rendition to justice program, the CIA used connections to foreign intelligence
services to coordinate the arrest of terrorists and hand them over to the FBI for
prosecution in U.S. courts.
292
Immediately after 9/11, states and their respective
agencies worked to improve law enforcement intelligence sharing. Through informal
288. JOHN R. WAGLEY, CONG. RESEARCH SERV., RL 33335, TRANSNATIONAL ORGANIZED CRIME:
PRINCIPAL THREATS AND U.S. RESPONSES 1112 (2006), available at http://www.fas.org/sgp/crs/
natsec/RL33335.pdf.
289. MALCOLM ANDERSON, POLICING THE WORLD: INTERPOL AND THE POLITICS OF
INTERNATIONAL POLICE CO-OPERATION 162 (1989). See also GILL, supra note 2, at 28 ([I]n recent years
law enforcement agencies have come to view serious crime as increasingly transnational and to exchange
information and techniques with foreign agencies.). See generally INTERNATIONAL POLICE
COOPERATION: A WORLD PERSPECTIVE (Daniel J. Koenig & Dilip K. Das eds., 2001).
290. Heymann, supra note 110, at 14 (noting that [p]olice agencies have a greater tradition of
sharing information across borders than intelligence agencies).
291. See RICHARD A. BEST, CONG. RESEARCH SERV., RL 30252, INTELLIGENCE AND LAW
ENFORCEMENT: COUNTERING TRANSNATIONAL THREATS TO THE US 1 (2001), available at http://www.
fas.org/irp/crs/RL30252.pdf.
292. MARGARET SATTERTHWAITE & ANGELINA FISHER, CTR. FOR HUMAN RIGHTS & GLOBAL
JUSTICE, N.Y. UNIV. SCH. OF LAW, BEYOND GUANTANAMO: TRANSFERS TO TORTURE ONE YEAR
AFTER RASUL V. BUSH 911 (2005), available at http://www.chrgj.org/docs/Beyond%20Guantanamo%20
Report%20FINAL.pdf. See also HUMAN RIGHTS WATCH, supra note 155, at 4 (During the decade prior
to 1998, the U.S. government used extraordinary rendition to bring 13 terrorist suspects to the United
States to stand trial on criminal charges.).
MOUs and formal agreements, states bolstered existing mechanisms of exchange,
such as Interpol and Europol.
293
States also concluded mutual legal assistance and
recognition treaties in order to alleviate some of the difficulties in arrest and
extradition of terrorist suspects.
294
Many experts still believe a more coordinated
law enforcement and intelligence approach would better combat both international
crime and terrorism.
295
As they did in the past, liberal democracies could harness
the connections of their intelligence agencies to help rather than hinder criminal
prosecutions. To encourage the arrest, prosecution, and conviction of perpetrators
of terrorism, intelligence should be preferably produced in a way that allows it to be
exploited as evidence, and law enforcement officials should use the tools at their
disposal, such as witness protection programs, to protect those intelligence sources
that might be endangered.
296

Aligning intelligence and law enforcement work is no easy task,
297
but the
conclusion that intelligence should automatically trump law enforcement is by no
means the logical one. Prosecution in civilian criminal courts undeniably exacts cost
and effort, requiring officials to assemble evidence, meet evidentiary thresholds, and
locate and arrest perpetrators abroad.
298
Nevertheless, most liberal democracies are
equipped with criminal statutes applicable to terrorist acts committed at home and
abroad. As a general matter, they boast successful records of prosecuting terrorist
acts and other transnational crimes.
299
Although some fear that court procedures will
reveal classified information and pose risks to intelligence sources and methods,
300

courts are adept at finding creative solutions to protect information while allowing
for an effective defense. Legislatures also can and have enacted statutes to protect
classified information while maintaining the integrity of criminal proceedings.
2. Benefits of Law Enforcement Intelligence and the Involvement of Courts
Though affected by some of the same problems as intelligence cooperation, law
enforcement networks benefit from greater transparency and the involvement of
courts. The involvement of courts has several distinct advantages. It advances
Western intelligences interests in accuracy and the pursuit of truth, which facilitates
293. WAGLEY, supra note 288, at 13 (discussing various mechanisms for exchange of criminal
information transnationally). See also Jorg Monar, Anti-Terrorism Law and Policy: The Case of the
European Union, in GLOBAL ANTI-TERRORISM LAW AND POLICY PAGE, 42530 (Victor V. Ramraj et
al. eds., 2005) (analyzing the law enforcement cooperation mechanisms adopted within the EU, including
Europol).
294. Council Directive 9153/03, Agreements on Extradition and on Mutual Legal Assistance, 2003
O.J. (L181).
295. WAGLEY, supra note 288, at 11.
297. Gregory F. Treverton, Terrorism, Intelligence and Law Enforcement: Learning the Right
Lessons, 18 INTELLIGENCE & NATL SECURITY 121, 122 (2003) (declaring a fundamental disconnect
between the intelligence culture and the law enforcement).
298. Bay, supra note 20, at 35657 (describing several difficulties with criminal prosecution).
299. Heymann, supra note 156, at 451 (noting the success of Department of Justice prosecutions for
terrorism); Richard Falk, Human Rights: A Descending Spiral, in HUMAN RIGHTS IN THE WAR ON
TERROR 225, 23637 (Richard Ashby Wilson ed. 2005); see also Mark A. Drumbl, Victimhood in our
Neighborhood: Terrorist Crime, Taliban Guilt, and the Asymmetries, 81 N.C. L. REV. 1, 73 (2002).
300. Bay, supra note 20, at 357.
the protection of democratic states, a task at the heart of these agencies missions. It
also tempers the impulse to rely on illiberal partners to extract information through
coercive tactics. Even if the safeguards to prohibit the use of coerced information
proposed in previous sections were adopted, one could imagine that an agency might
still receive, use, and forward intelligence that has been obtained in violation of
human rights law.
301
The best way to reduce this potential is to facilitate challenges to
the information. Criminal trials afford the affected individual the means to challenge
the credibility, reliability, and accuracy of information. The rights to present a
defense, to confront ones accusers, to protect against self-incrimination, and to
assistance of counsel reduce the risk of error and incidence of government
wrongdoing.
Anecdotal evidence from post-9/11 counterterrorism efforts indicates the
importance of these procedural safeguards. A U.K. court review of the detention of
several individuals found that MI5 had relied on very weak evidence, and falsely
claimed that weapons had been found in order to detain them.
302
The courts
involvement also enhanced MI5s accuracy. After the proceedings revealed that the
purpose behind a group of Muslim mens visit to Dorset had not been to elect a
terrorist leader [as MI5 had asserted] but to get away from their wives for the
weekend, MI5 withdrew its assessment of their dangerousness.
303

In the United States, courts played a similar role. In one case, intelligence
officers and prosecutors were willing to accept intelligence and witnesses provided by
a Russian agency indicating a U.S. citizen was a well-known terrorist financier. The
information turned out to be utterly unreliable as a case of false identity.
304
Yet only
because the individual concerned was prosecuted in a court of law did the falsity of
this information come to light.
Court proceedings lessen the risk of Western involvement in or reliance on
coerced evidence, inherent in cooperation with less-reputable partners, by exposing
wrongdoing and improprieties. For example, a U.S. court uncovered allegations of
torture when Wang Zong Xiao, a defendant arrested and charged in China, came to
the United States to testify against other defendants and recanted his confession,
claiming coercion.
305
Similarly in Britain, an MI5 officer admitted in court that MI5
possibly assessed evidence obtained through torture as reliable and that
administrative agencies likely relied on that evidence.
306
By creating incentives for
Western intelligence to demand high professional and ethical standards from their
intelligence partners, participation of courts may generate more credible, efficient
intelligence exchange against transnational threats.
Court proceedings help ensure that the right person is subject to sanctions for
wrongdoing. Interrogation by intelligence officials can subject individuals to the
302. Mark Phythian, Still a Matter of Trust: Post-9/11 British Intelligence and Political Culture, 18
INTL J. OF INTELLIGENCE AND COUNTERINTELLIGENCE 653, 66970 (2005). Among the intelligence that
MI5 relied upon was the observation that the detainees were excessively security conscious while
shopping. Id.
303. Id. at 669.
304. Id.
305. Amann, supra note 68, at 55758 (The desire to proceed with the joint effort seemed to have
blinded the United States prosecutor to earlier indications that the confession might have been coerced
and unreliable.).
306. AMNESTY INTERNATIONAL, supra note 145, at 23.
impossible situation of having to disprove an allegation, in many countries under the
threat of torture. By contrast, criminal procedures place the burden on the
government to make a case, permit the defendant to engage in discovery, and require
disclosure of exculpatory information known to or in the possession of the
government. Whereas the current transnational system which errs by not
prosecuting those who have committed violent acts and by detaining innocent people
who have not, criminal procedures aim to convict the guilty and release the innocent,
in the best interest of Western states and their intelligence agencies.
Accuracygetting the right guyhas a better outcome both in terms of the
reputation of liberal democracies and of the promotion of the rule of law in our
partner nations. The involvement of courts bestows counter-terrorism efforts and
the agencies that implement them with greater legitimacy and public support. By
upholding the rule of law, trials give the countries involved moral legitimacy and
encourage the development of international legal norms to deal with terrorism.
307

Their capacity for verification portends some measure of quality control, which could
remedy certain irrationalities in current intelligence gathering and sharing
arrangements.
Although this discussion has focused largely on the use of courts independent of
the executive, certain administrative procedures might also be developed to permit
challenges to and correction of inaccurate intelligence information and to trigger
investigation of possible wrongdoing or statutory violations through intelligence
networks.
308
Some states have already developed an administrative right for redress
of abuses produced through improper intelligence gathering. The United Kingdom
instituted a tribunal where a person can file complaints regarding actions taken by or
on behalf of the intelligences services with regards to him, his property, or
communications.
309
Similarly in Australia, a Security Appeals Tribunal, which has a
high-level of security clearance, provides a forum for complaints of illegal
intelligence activity.
310
In addition, an Inspector-General of Intelligence and Security
acts as an ombudsman for the public and has the power to demand information,
examine the records of departments, and investigate complaints about the propriety
and accuracy of information gathered.
311

International tribunals present another intriguing alternative. Some contend
that trials before an ethnically and nationally diverse panel of judges offer the most
appropriate forum for criminal prosecution of transnational crimes.
312
International
tribunals have the comparative advantage of involving the global public,
standardizing cooperation and punishment, and enhancing legitimacy of
307. Bay, supra note 20, at 35556.
308. See Intelligence Bill of Rights, supra note 189, para. 10 (advising that [c]itizens and resident
aliens who believe that their rights were violated by intelligence agencies . . . be able to apply to the
intelligence agency concerned for access to information held by the state and or such agency about
them).
309. BIRKINSHAW, supra note 69, at 47.
310. Id. at 54.
311. Id. at 6061, 242.
312. Drumbl, supra note 299, at 2; see generally Laura A. Dickinson, Using Legal Process to Fight
Terrorism: Detentions, Military Commissions, International Tribunals, and the Rule of Law, 75 S. CAL. L.
REV. 1407 (2002) (arguing in favor of an international process for terrorism prosecutions).
counterterrorism efforts and factual narratives developed at trial.
313
Laura Dickinson
argues that international trials could help to strengthen the needed intelligence-
sharing networks and could help to provide a framework for screening sensitive
information that would have greater legitimacy than a purely U.S.-run process.
Accordingly, Muslim states might more willingly extradite suspects and provide full
cooperation.
314

Whether national or international, judicial or administrative, court proceedings
conducted with respect for due process and individual rights should diminish the
affinity of many civilians for terrorist groups and strengthen resolve against such
groups. The criminal process establishes a historical record of events and exposes
evidence of wrongdoing, which can serve to dissipate terrorists cult status.
315
Perhaps
the greatest value is that [l]aw enforcement, with its focus on the illegal act itself,
removes the temptation to try to judge between just and unjust motivations,
legitimate and illegitimate concessions, worthy and unworthy political causes.
316

CONCLUSION
If anything is certain in the murky world of intelligence collection, it is that
intelligence sharing will continue. Though the partners and the level of sharing may
change, intelligence sharing will remain an indispensable component of any effective
strategy against transnational threats. But, as this Article has shown, intelligence
sharing networks must not remain as they are. As they function now, these networks
strain the principles of democracy and accountability. They create ample
opportunities for even the most professionalized agency to subvert international and
national legal restraints through collusion with network partners, and resulting in
untold human rights abuses. More fundamentally, the operation of intelligence
sharing networks challenges our democratic system, eviscerating notions of national
jurisdiction and traditional mechanisms of managing intelligence agencies.
Reconciling the necessity of transnational cooperation and the constraints of
liberal democracy should therefore be a top priority for national executives,
legislatures, and international bodies. Network sanctions and high professional
standards, although desirable, should not be the only means to evaluate and improve
intelligence cooperation. National democratic bodies, international institutions, and
regional alliances all have a role to play. As implausible as it may seem, executive
and legislative engagement with their counterparts in allied states may prove
necessary to ensure intelligence cooperation that is effective and furthers the larger
interests of Western democracies. A return to law enforcement and use of the
judiciary is essential for the accuracy and legitimacy of efforts to counter
transnational threats.
Without real concerted change, we will continue to see the proliferation of
inaccurate information, brutal tactics, and intelligence failures due to improper
information sharing. Accountability and democracy will continue to suffer. And the
network arrangements of our intelligence agencies will continue to challenge our
313. Drumbl, supra note 299, at 1314.
314. Dickinson, supra note 312, at 1448.
315. Falk, supra note 299, at 23738.
316. United States Institute of Peace, The Diplomacy of Counterterrorism: Lessons Learned,
Ignored, and Disputed 56 (Jan. 14, 2002), available at http://www.usip.org/files/resources/sr80.pdf.
conceptions of international law, accountability, and ultimately the role of our own
state in the world.

Democracy, Human Rights and Intelligence Sharing

Uploaded by

Copyright:

Available Formats

You might also like

Democracy, Human Rights and Intelligence Sharing

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Democracy, Human Rights and Intelligence Sharing

Uploaded by

Copyright:

Available Formats

09 Sepper PUB_FINAL1 (Do Not Delete) 11/21/2010 5:33 PM

You might also like